Large and small particle size screening of organic compounds in urban air

Atmospheric Environment Vol. 27B. No. 2, pp, 251 263, 1993.

0957 1272/93 $6.00+0.00 © 1993 Pergamon Press Ltd

Printed in Great Britain.

LARGE A N D SMALL PARTICLE SIZE S C R E E N I N G O F O R G A N I C C O M P O U N D S IN URBAN AIR MERCE ACEVES Environmental Control Service, Entitat Metropolitana de Barcelona, Carrer 62, 420, Edifici A Zona Franca, 08004-Barcelona, Catalonia, Spain

and JOAN O. GRIMALT* Department of Environmental Chemistry (C.I.D.-C.S.I.C.),Jordi Girona, 18, 08034-Barcelona, Catalonia, Spain (First received 4 June 1992 and in final form 16 November 1992)

Abstraet--A gas chromatographic (GC)-mass spectrometric (MS) study of the large ( > 7.2 #m) and small (<0.5#m) particle fractions of aerosols collected in Barcelona (Catalonia, Spain) has allowed the identification of the major sources of airborne organic matter in urban environments. The extracts have been column chromatography separated in three fractions encompassing aliphatic hydrocarbons, aromatic hydrocarbons + semi-polar compounds, and polar products. The whole analytical procedure has allowed the identification of 186 molecular species that have been grouped according to their precursors. The study has also afforded the characterization of organic source-particle size associations as well as the effects of seasonality on the occurrence of the compounds identified. Key word index: Urban aerosols, particle-sized atmospheric particulates, hydrocarbons, alcohols, acyl

glycerols.

INTRODUCTION The emission of pollutants to the atmosphere often involves the direct uncontrolled exposure of large populations to toxic substances. This is especially relevant in urban areas where the proximity between humans and pollutant sources is closest. Thus, a direct cause-effect relationship between the atmospheric release of allergenic products and urban asthma outbreaks has recently been documented (Anto et al., 1989a, b). In addition, many organic compounds found in urban air are mutagens or carcinogens (NRC, 1983; Guicherit and Schulting, 1985; Tancrede et al., 1987). The interrelations between particle-size and atmospheric transport of organic compounds constitutes one of the aspects less understood. The distribution of n-alkanes and polycyclic aromatic hydrocarbons (PAH) among particles of different sizes has sometimes been described (Pierce and Katz, 1975; Van Vaeck and Van Cauwenberghe, 1978, 1985; Miguel and Friedlander, 1978; Katz and Chan, 1980; Sicre et al., 1987; Pistikopoulos et al., 1990) but environmental effects such as seasonality have been rarely considered. Furthermore, integrated studies on the whole solvent extractable species still remain to be * Author to whom correspondence should be addressed.

addressed. However, the atmospheric behaviour of the more polar products in urban areas may have significant health implications (Aceves et al., 1991). Since 1985 we have regularly monitored the composition of solvent extractable compounds in the aerosol of Barcelona (Catalonia, Spain) using methods based on gas chromatography (GC) and GC coupled to mass spectrometry (GC-MS). This regular sampling and analysis has afforded a detailed knowledge on the distributions of the major lipids among the atmospheric particulates of the city. They exhibit seasonal variations that reflect the changes in predominant pollution and natural sources along the year. Aerosol samples corresponding to these well-defined situations have been collected using a particle-size impactor. This fractionation has resulted in the decoupling of the previous GC profiles in a series of standard GC patterns that illustrate different processes of incorporation of the organic compounds to the urban atmosphere. The standard GC traces corresponding to the central area of Barcelona, the most densely populated area, are described in the present paper. Their composition is interpreted in terms of predominant input sources and compound-particle associations. Since many of the pollutant products identified are related to traffic exhaust and fossil fuel burning, the seasonally dependent GC distributions reported here can be used 251

252

M. ACFVESand J. O. GRIMAL'r

as reference for c o m p a r i s o n with those obtained in other cities.

EXPERIMENTAL

Materials Pestipur-grade dichloromethane was purchased from Mallinckrodt (Paris, U.S.A). Chromatography quality nhexane, methanol, isooctane, neutral silica gel (Kieselgel 40, 70-230 mesh) and alumina (aluminum oxide 90 actiye, 70-230 mesh) were from Merck (Darmstadt, Germany). The Soxhlet apparatus for 24 h. After solvent evaporation, the Germany). The glass microfibre filters were purchased from Whatman (Maidstone, U.K.). The silica gel, the alumina and the Soxhlet cartridges were extracted with dichloromethane-methanol (2: 1, v/v) in a Soxhlet apparatus for 25 h. After solvent evaporation, the silica and the alumina were heated for 12 h at 120 and 350°C, respectively. A total of 5 % (w/w) of Milli-Q-grade water was then added to the chromatographic adsorbents for deactivation. The glass-fibre filters were kiln-fired for 12 h at 400°C and weighed prior to sampling. The purity of the solvents was checked by concentrating under vacuum 100 ml of solvent to 10 pl for GC analysis. Blank requirements were as follows: splitless injection of 2/ll should result in chromatograms with no unresolved GC envelope and only a few peaks, representing up to 1 ng in terms of their flame ionization detector response. This threshold, under the above dilution factor, is equivalent to less than 0.01 pgm 3 when referred to 100 ml of solvent used for the extraction of one-third of a filter that corresponds to 2000 m 3 of air sample.

Sampling, extraction and fractionation The air samples were taken with a High-Vol pumping system (CAV-P; MCV, Collbato, Catalonia, Spain) equipped with a cascade impactor (Sierra-Andersen, model 235). The sampling periods were 48 h at 60 m 3 h - 1. After sampling, the filters were stored at - 20° C until analysis in the laboratory. One-third of each filter was Soxhlet-extracted with 100 ml of dichloromethane for 24 h. The extract was vacuum and nitrogen evaporated until almost dryness and diluted to 0.5 ml with n-hexane. Then, it was fractionated by column chromatography according to previously established methods (Aceves et al., 1988). These methods were scaled down to suitable amounts of solvents and packings according to the small lipid content of the filters. A column filled with 1 g each of 5% water-deactivated alumina (top) and silica (bottom) was used. The aliphatic hydrocarbons were obtained in the first fraction (4 ml of n-hexane), the aromatic hydrocarbons were collected in the second fraction (4 ml of 20% dichloromethane in n-hexane), and the polar compounds eluted in the third fraction (4 ml 20% methanol in dichloromethane). These fractions were vacuum and nitrogen concentrated to almost dryness and redissolved with isooctane. The recovery factors of this fractionation process were in the order of 90-100% for aliphatic and aromatic hydrocarbons and sterols.

program from 60 to 300'~C at 6"C min 5, injector and detector temperatures of 280 and 330°C, respectively, and hydrogen as carrier gas (50 cms-1). The oven temperature program for the GC-MS analyses was from 60 to 280°C at 6~'C min- 1, injector and transfer line temperatures were 280 and 300°C, respectively, and helium was the carrier gas (50cms-a). Data were acquired in the EI mode (70 eV), scanning from 40 to 600 mass units at 1 s per decade. In both cases the injector was in the splitless mode (1 /~1,hot needle technique), the split valve being closed for 35 s.

Identification and quantitation Compound identification was based on the GC-MS data and on coinjection with authentic standards. The quantitations were performed from the GC profiles using the external standard method. Samples and standards were repeatedly injected until less than 5% dispersion in the area measurements was observed. An external standard containing n-eicosane, n-tricosane, n-octacosane, n-dotriacontane and n-hexatriacontane was used for the alkane fraction; the aromatic hydrocarbons were quantitated with a standard containing phenanthrene, fluoranthene, pyrene, benzo (ghi) fluoranthene, benzo(a) anthracene, chrysene, benzo (e) pyrene, dibenzo(a,h) anthracene, benzo(ghi) perylene and coronene. These two standards were also used as testmixtures for the assessment of undesired column adsorption effects or sample discrimination in the injection. The polars were quantitated with a standard containing coprostanol and n-octacosan- 1-ol.

RESULTS AND DISCUSSION T h e sampling site was located in the Barcelona city centre, in a square (Plaga Molina) undergoing heavy traffic. The high-volume p u m p was situated at 3 m above ground level a n d 1 0 0 m from any vertical obstacle. The m a x i m u m t e m p e r a t u r e during s u m m e r collection was 26°C while in winter it was 15°C. Both sampling periods were performed in conditions of atmospheric stability, in the absence of precipitation a n d with a wind speed less t h a n 4 m s - 1 at 25 m a b o v e g r o u n d level. The a m b i e n t air entered into the high-volume shelter a n d cascaded t h r o u g h the succeeding orifice stages undergoing progressively higher linear velocities. Thus, the particles were inertially impacted o n t o the collection glass-fibre sheets situated in each stage, the coarser being retained in the first stages a n d the finer in the latter. The smallest particles were collected with a back glass-fibre filter. The c o m p o s i t i o n of the first sheet a n d the final filter are c o m p a r e d in this study. They respectively correspond to cut-off diameters of > 7.2 and < 0.5 pm, with a collection efficiency of 50%.

Instrumental analysis

Aliphatic hydrocarbons

The samples were analysed by gas chromatography (GC) and GC coupled to mass spectrometry (GC-MS). These analyses were performed with a Carlo-Erba 5300 equipped with a flame ionization detector, and with a HewlettPackard 5970 provided with a data aquisition system HP5994A, respectively. 25 m x 0.2 mm i.d. HP-5 (film thickness 0 . l l # m ) (Hewlett-Packard, Palo alto, CA) and 30m x 0.25 mm i.d. DB-5 (film thickness 0.2 #m) (J&W Scientific, Folsom, CA) fused-silica capillary columns were used for the analysis of the hydrocarbon and polar fractions, respectively. The GC analyses were performed with an oven temperature

The G C profile of the coarse winter particles (Fig. 1 a n d Table 1) is d o m i n a t e d by n-alkane distributions ranging between Ct6 a n d Ca6 without o d d - e v e n carb o n n u m b e r predominance. O t h e r characteristic resolved peaks in this G C trace encompass regular isoprenoids, pristane (No. 4) a n d p h y t a n e (No. 6), a n d olephines, n-hentriacontene (No. 23) and n-dotriacontene (No. 27). These c o m p o u n d s overlay a n unresolved complex mixture ( U C M ) of h y d r o c a r b o n s eluting

Particle size screening of organic compounds Summer Dp > T..2/zm

Winter

Dp >

7.2#m

,8

253

19 24

192o2428 1516 ~2.5

9 io

15 30

I III IJLJJ; I Ii o. II 12 Ii [

24

Winter Dp < 0.Sp.m

Summer Dp < 0.5#m

23

25

'27

20 t~4 19/I 12t 1415~130

8

" ' ~ 32 12

J

II I0

3

I 12

23578 I ~

1

20

I

24

1

28

I

32

I

36

I

40

I

44

I

48

L

16

I

20

9 l

24

l

28

l

32

l

36

l

40

I

44

1

48

I

52

Fig. 1. GC profiles of the aliphatic hydrocarbons present in the particle-fractionated aerosols collected in Barcelona. The numbered peaks refer to Table I.

over the whole temperature range of the chromatogram. The n-alkanes without odd-to-even carbon number predominance occurring together with pristane and phytane, and an unresolved mixture of aliphatic hydrocarbons are characteristic of petroleum residues and vehicular exhausts (Simoneit, I984; Hampton et al., 1983). Similarly, the alkenes are likely pyrolysis products in the vehicular combustion processes. Thus, pyrolysis naphtha produced by thermal cracking of naphtha contains large quantities of alkenes and aromatics (Albaiges and Guardino, 1980). The winter samples exhibit aliphatic hydrocarbon profiles containing a higher proportion of UC M in the finer than in the coarser particles. Their n-alkane distributions range between C16 and C36 an(t show no odd/even carbon number preference. Other major resolved peaks in these fractions correspond to C27---C34 17~(H), 21fl(H)-hopanes which is consistent with the petroleum-related origin of these hydrocarbon mixtures (Simoneit, 1984; Farrington and Meyers, 1975). In fact, the whole chromatographic trace is a close match of the aliphatic hydrocarbons resulting from combined exhaust particulates of gasoline and diesel powered vehicles, the portion of UCM eluting at a higher retention time and the 17~(H), 21fl(H)hopanes corresponding to unburned lubricating oil residues (Trier et al., 1987, 1990). This lubricating oil contribution may originate from leakages in the valve stem seals or piston rings of the engines (Furuhama and Hiruma, 1979).

The summer samples exhibit two major contrasting features with respect to the winter aerosols: first, the low proportion of compounds eluting before C25, which is observed for both the straight chain and the UCM hydrocarbons and corresponds to a volatilization effect due to the higher ambient temperature; second, the higher input of biogenic material that is reflected in the composition of the odd carbon numbered C 25-C33 n-alkanes in the coarse particles. These n-alkanes are characteristic of higher plant wax sources (Eglinton and Hamilton, 1967) and they preferentially occur in the larger size fractions (Sicre et al., 1987). The quantitative values corresponding to the major peaks identified in Fig. 1 are listed in Table 1. The comparison of the concentrations between coarse and fine particles shows an approximate ratio of 1 : 10 for both resolved peaks and UCM. This trend is consistent with previous observations showing that the petrogenic residues preferentially occur in the submicron particle fraction (Pierce and Katz, 1975; Van Vaeck and Van Cauwenberghe, 1978; Katz and Chan, 1980; Sicre et al., 1987). The comparison between winter and summer samples shows a higher amount of petrogenic hydrocarbons in the cold season. This difference likely corresponds to a higher use of petrogenic fuels in winter although volatilization, and possibly photooxidation, contribute to hydrocarbon decrease in summer. The ratio between resolved compounds and UCM usually ranges between 7 and 15

M. ACEVESand J. O. GRIMALT

254

Table 1. Aliphatic hydrocarbons identified in the particulate fractions of the aerosol samples collected in Barcelona City* No.

1

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

Compound

Diagnostic ions

n-Pentadecane n-Hexadecane n-Heptadecane Pristane n-Octadecane Pbytane n-Nonadecane n-Eicosane n-Heneicosane n-Docosane n-Tricosane n-Tetracosane n-Pentacosane n-Hexacosane n-Heptacosane n-Octacosane 18ct(H)-22,29,30-Trisnorhopane 17ct(H)-22,29,30-Trisnorhopane n-Nonacosane n-Triacontane 17ct(H),21 fl(H)-30-Norhopane

212 226 240 268 254 282 268 282 296 310 324 338 352 366 380 394 l 91 191 408 422 191

Winter aerosol Coarser Fine S

Summer aerosol Coarse Fine

0.17

1.3

0.04

0.46

0.43

2.0

0.03

0.48

0.90 1.5 1.6 1.6 1.5 1.3 1.1 1.1 1.4 1.4

3.3 5.7 8.9 16 16 16 16 16 15 4.5

0.02 0.04 0.04 0.08 0.15 0.19 0.41 0.35 0.77 0.35

0.50 0.61 1.2 2.7 5.5 8.5 11 8.2 8.9 5.1

2.1 2.0

10 3.5

1.2 0.54

9.0 4.5

2.6

12

1.4

3-Methyl-17ct(H),21fl(H)-30norhopane n-Hentriacontene n-Hentriacontane 17ct(H),21 fl(H)-Hopane 3-Methyl- 17~(H),21 fl(H)-hopane n-Dotriacontene n-Dotriacontane 17ct(H),21 fl(H)-Homohopanes n-Tritriacontane 17~(H),2 lfl(H)-Bishomohopanes n-Tetratriacontane 17a(H),21fl(H)-Trishomobopanes n-Pentatriacontane

191 434 436 191 191 448 450 191 464 191 478 191 492

18

2.1

3.2

0.55

5.0

1.8

4.2

0.71

1.3

0.97

0.45

1.8

1.3

0.98

0.40

1.7

7.6 18

110 760

11

17a(H),21fl(H)Tetraquishomohopanes n-Hexatriacontane

Total n-alkanes Unresolved complex mixture

191 506 27 210

160 2400

*Concentrations in ng m- 3. Only the major compounds have been quantitated, t > 7.2/~m. ++<0.5 pm.

with the exception of the coarse summer particles in which it is 2.4. This latter case reflects the contributions of biogenic sources that are represented by the above-mentioned distributions of n-alkanes.

Aromatic hydrocarbons and other semi-polar compounds including lipids Polycyclic aromatic hydrocarbons. The G C traces of the aromatic hydrocarbon fractions (Fig. 2 and Table 2) are dominated by parent PAH, namely fluoranthene (No. 54), pyrene (No. 58), benzo(ghi)fluoranthene (No. 77) 4(H)-cyclopenteno(cd)pyrene (No. 81), benz(a)anthracene (No. 82), chrysene/triphenylene (No. 83), benzofluoranthenes (Nos 94-96), benzopyrenes (Nos 97, 98), indeno[1,2,3-cdlpyrene (No. 110), benzo(gh0perylene (No. 114) and coronene (No. 128).

The abundance of catacondensed structures and the predominance of parent hydrocarbons over alkylated homologues indicate that these distributions of resolved compounds are of pyrolytic origin (Simoneit, 1985; L a F l a m m e and Hites, 1978). Further assessment into processes determining the composition of these hydrocarbons can be obtained from the consideration of some concentration ratios. Thus, the fluoranthene/(fluoranthene +pyrene) ratio is rather uniform, 0.40 and 0.46, approaching that reported for exhausts of gasoline-fueled vehicles (Grimmer and Hildebrandt, 1975). The indeno [1,2,3cd]pyrene/(indeno[1,2,3-cd]pyrene + benzo(ghi) perylene) ratio has values between 0.44 and 0.52 in winter and about 0.32 in summer. The first corresponds to oil- (Baek et al., 1991; Grimmer et al., 1980) or coal-

Particle size screening of organic compounds

255

Summer Dp > 72~m

Winter Dp > 7..2p.m se

4~ 10611~.,120

20 | I~ ,,e/,~ ,od,-II 1,24

I01

,%

'

~

94 ~

5"7

,~,~

,.Jl j

Summer Dp < 0.5~m

Winter Dp < 0.SFm

58

114 82

99

94

I~

128 6772 I

I 8

I 12

I 16"

I 20

58~

1 ;=4

I

I

I

28

32

36

I 40

I 44

1 12

I 16

I 20

I 24

I 28

I 32

I 36

I

4O

1

44

I

48

Fig. 2. GC profiles of the aromatic hydrocarbons and semi-polar compounds present in the particle-fractionated aerosols collected in Barcelona. The numbered peaks refer to Table 2.

(Grimmer et al., 1983; Ratajczak et al., 1984) fired residential furnaces and the second to diesel oil vehicle exhausts (Grimmer, 1986, pers. commun.). This change between predominant winter and summer ratio reflects the seasonality of some urban activities such as domestic heating. About 25% of the central heating systems in the city are oil or coal powered and the rest use natural gas (Ajuntament de Barcelona, 1985). The benzo(e)pyrene/(benzo(e)pyrene + benzo(a)pyrene) ratio is also seasonally dependent, with values of 0.47 in winter and 0.69 in summer. The former value is characteristic of oil- and coal-fired domestic heating emissions (Grimmer et al., 1980, 1985) and the increased summer ratio may partially reflect the lack of domestic heating. However, a 0.69 value is too high to be attributed to vehicular traffic emissions (Grimmer and Hildebrandt, 1975). This high ratio is in fact due to the preferential degradation of benzo(a)pyrene with respect to benzo(e)pyrene after emission to the atmosphere (Nielsen, 1988), a process that is enhanced by the high summer temperatures. The fast decay of some specific PAH during this season is also reflected in the strong decrease of benz(a)anthracene with respect to chrysene + triphenylene and the disappearance of 4(H)-cyclopenteno(cd)pyrene (Nielsen, 1988; Kamens et al., 1988). This

latter compound is characteristic of gasoline-fueled vehicles (Greenberg et al., 1981; Grimmer et al., 1980) and is only found in winter samples. A similar seasonal trend has been observed in a study performed in New Jersey (Greenberg, 1989). The total concentrations of these airborne PAH are more than 50 times higher in the fine than in the coarse particle fraction. This trend parallels the preferential association of the anthropogenic n-alkanes to the finer aerosols and essentially reflects the gas-particle conversion of the heavier PAH (molecular weight > 202) in the submicron range after emission to the atmosphere (Pilinis et al., 1987). Conversely, the smaller PAH (molecular weight < 202) remain in the gas phase being adsorbed on pre-existing particles upon cooling (Pistikopoulos et al., 1990). This latter mechanism gives rise to a higher proportion of the lower molecular weight PAH in the coarser fractions. The same mechanism is likely to be responsible for the occurrence of PAH with molecular weight ranges between phenanthrene and pyrene in urban street dusts (Takada et al., 1990). The total amount of PAH is about 14 times higher in winter than in summer. This is consistent with previous studies (Pierce and Katz, 1975; Van Vaeck and Van Cauwenberghe, 1985; Pistikopoulos et al., 1990) and reflects the wider occurrence of combustion

256

M. ACEVESand J. O. GRIMALT

processes during the cold season as well as the higher association of these hydrocarbons to the atmospheric particles at lower ambient temperatures. Accordingly, the GC traces of the summer aerosols exhibit a strong decrease of the more volatile PAH (molecular weight < 226) both in the coarse and in the fine particle fractions. An equivalent decrease is also observed among the n-alkanes. In addition to these resolved compounds, an important UCM is found in all the GC profiles of these second column chromatography fractions. This mixture is related to petrogenic sources as that observed in the aliphatic hydrocarbon chromatograms. In fact, the "baseline humps" from the GC traces of Figs 1-3 merely reflect the fractionation of the same UCM by the column chromatography method selected for this study. These petrogenic inputs are also represented by some minor resolved compounds such as the alkylated phenanthrenes, dibenzothiophenes and chrysenes/triphenylenes.

Biogenic products. In these semi-polar fractions the terrestrial vegetation contributions are represented by lipid distributions such as odd carbon numbered C29-C33 n-alkan-2-ones (Grimalt and Albaiges, 1990), and even carbon numbered C26-C32 n-aldehydes (Prahl and Pinto, 1987), C30-C36 dimethyl n-alkanedioates (Cranwell, 1977) and C32-Cas saturated and unsaturated wax esters (Kolattukudy, 1970). These lipids are essentially found in the coarse particles collected in summer. In fact, the occurrence of naldehydes, and to a lower extent n-alkan-2-ones, is a major qualitative feature that allows differentiation between sampling seasons. These trends are in agreement with the above described occurrence of biogenic n-alkanes and reflect the size range of the particles that are usually emitted from biogenic sources. In this respect, the occurrence of retene in the summer coarse particles points to an origin for this hydrocarbons that, instead of wood combustion processes (Ramdahl, 1983a), is also related to higher plant

Table 2. Aromatic hydrocarbons and semi-polar compounds identified in the particulate fractions of the aerosol samples collected in Barcelona* No.

Compound

37 38 39 40 41 42 43 44 45 46 47 48

Fluorene Dibenzothiophene Phenanthrene Anthracene 3-Methylphenanthrene 2-Methylphenanthrene 2-Methylanthracene

49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73

4H-Cyclopenta(def)phenanthrene 4/9.-Met hylphenanthrene 1-Methylphenanthrene Methyl hexadecanoate 4H-Methylcyclopenta (def) phenanthrene 2,7-Dimethylphenanthrene Dimethyl-178 Dimethyl-178 Dimethyl-178 Dimethyl-178 Fluoranthene Trimethyldibenzothiophene Acephenanthrylene Trimethyldibenzothiophene Pyrene

Diagnostic ions 166 184 178 178 192 192 192 190 192 192 270 204 206 206 206 206 206 202 226 202 226 202

4H-Dimethylcyclopenta(def)phenanthrene Trimethyldibenzothiophene Methyl octadecanoate Trimethyl-178 Trimethyl-178 Trimethyl-178 Trimethyl-178 8/7-Methylpyrene Benzo(a)fluorene Retene Benzo(b)fluorene 2-Methylpyrene 4-Methylpyrene 1-Methylpyrene Dimethyl-202

218 226 298 220 220 220 220 216 216 234 216 216 216 216 230

Winter Summer aerosol aerosol Coarser Fine:~ Coarse Fine

0.24 0.02

2.6 0.96

0.028 nd

0.11 nd

0.89

10

0.036

0.57

1.1

15

0.043

0.83

Particle size screening of organic compounds

257

Table 2. (Continued) No.

Compound

74 75 76

Methyl dehydroabietate Dimethyl-202 Benzo(b)naphtho[2,1-d]thiophene

77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 1l 5

Benzo(ghi)fluoranthene

116 117 118

119 120 121

122 123 124 125 126

127 128 129 130 131 132 133 134

Benzo(c)phenanthrene Benzo(b)naphtho[ 1,2-d]thiophene Benzo(b)naphtho[2,3-d]thiophene 4~H)-Cyclopenteno(cd)pyrene Benz(a)anthracene Chrysene/triphenylene Methyl-234 Methyl-234 Methyl-234 n-Octyl hexadecanoate Methylchrysene Methyl-228 Methyl-226 Methyl-226 Binaphthyl n-Octyl octadecanoate

Benzo(b/j)fluoranthene Benzo(k)fluoranthene Benzo(a)fluoranthene Benzo(e)pyrene Benzo(a)pyrene Squalene Perylene n-Hexacosanal Methyl-252 Quaterphenyl Methyl-252 Methylene-252 n-Octacosanal Dibenzonaphthothiophene

Indeno[7,1,2,3-cdefJchrysene n-Nonacosan-2-one Indeno[ 1,2,3-cd]pyrene Dibenzo(aj)anthracene Dibenzo(ah)anthracene Benzo(b)chrysene

Benzo(ghi)perylene Anthantrene n-Triacontanal n-Hentriacontan-2-one n-Hexadecyl hexadecenoate Dimethyl triacontanodioate n-Hexadecyl hexadecanoate n-Dotriacontanal n-Octadecyl hexadecenoate Dimethyl dotriacontanodioate n-Octadecyl hexadecanoate Dibenzopyrene n-Tritriacontan-2-one Dibenzopyrene Coronene n-Eicosyl hexadecenoate Dimethyl tetratriacontanodioate n-Eicosyl hexadecanoate n-Docosyl hexadecenoate Dimethyl hexatriacontanedioate n-Docosyl hexadecanoate

Diagnostic ions

Winter aerosol Coarser Fines

Sulnmer aerosol Coarse Fine

239 230 234 226 228 234 234 226 228 228 248 248 248 112 242 242 240 240 254 112 252 252 252 252 252 410 252 380 266 306 266 264 408 284 276 422 276 278 278 278 276 276 436 450 478 98 480 464 506 98 508 302 478 302 300 534 98 536 562 98 564

Total HAP Unresolved aromatic hydrocarbons *Concentrations in ng m - a. t > 7.2/am. :~< 0.5 #m.

0.13

12

nd

1.2

0.06 0.06 0.13

23 8.8 16

nd nd 0.040

nd nd 1.7

0.10

27

nd

2.6

0.08 0.09

19 22

nd nd

1.7 0.76

0.11

24

0.018

1.5

0.10

29

0.020

3.2

0.06

10

nd

0.95

4.1

97

220 1800

0.18 36

16 270

258

M. ACEVESand J. O. Winter Dp > 7,2#m

144

159

GRIMALT

mmer i

176

Dp > 7.2#m

137~,t

t75

137

181

17`21"/51 I

18P

~44

~

I 176

t53 ~

146 149

177 I ~ ' 1~6

/

i~o9 141

Winter Dp < O,5~m

153 1159

182

1?'2

Summer Dp < 0.5/.¢m t72

15"7

137.

163

17~

186

J I 8

I

I

I2

16

20

L

1

I

24

28

32

36

1 40

I 44

,I 8

l 12

I 16

I 20

L 24

J 28

1 32

i 36

I 40

1 44

Fig. 3. GC profiles of the polar compounds present in the particle-fractionated aerosols collected in Barcelona. The numbered peaks refer to Table 3.

inputs, namely coniferous vegetation (Simoneit and Mazurek, 1982). Squalene (No. 99) is another biogenic product very abundant in some particle fractions that is usually representative of microorganism contributions (Bird and Lynch, 1974). Polar compounds

The major compounds in the GC profiles of the polar fractions are phthalates, fatty acids and alcohols. Phthalates are in fact the predominant constituents in all cases. These compounds are found world-wide in soils (Russel and McDuflie, 1983) and waters (Ritsema et al., 1989) reflecting the widespread use of plasticizers. Di-n-octyl phthalate (No. 159) is the major homologue, its presence is concurrent with other n-octyl esters identified in these samples such as di-n-octyl hexanedioate (No. 153), a major compound in fraction 3, and n-octyl hexadecanoate (No. 87) and octadecanoate (No. 93), in fraction 2. Alcohols. n-Alkan-l-ols and sterols constitute the second major group of polar compounds. The nalkan-l-ols encompass a distribution of even carbon numbered homologues ranging between C16 and C32 in which the C2s C32 mode is predominant. This type of distribution is characteristic of contributions from higher plant waxes (Eglinton and Hamilton, 1967), its occurrence is consistent with the above-described

distributions of aldehydes, ketones, wax esters and dicarboxylic acids found in the semi-polar fraction. These alcohols are seasonally and particle-size dependent. The higher amount of n-alkan-l-ols in summer corresponds to a higher emission by the terrestrial vegetation in the warm seasons, a trend that we have repeatedly observed in studies of aerosols from rural areas. The composition of these linear alcohols is rather independent of aerosol size but their relative proportion is higher in the particles > 7.2/~m which is consistent with the above-described preferential distribution of biogenic products in the coarser fractions. The sterols constitute a mixture of C27-C29 compounds predominated by 24-ethylcholest-5-en-3fl-ol (No. 182) in which cholest-5-en-3fl-ol (No. 176), 24methylcholest-5-en-3fl-ol (No. 179) and 24-ethylcholest-5,22-dien-3fl-ol (No. 180) are major species. This composition is also representative of inputs from higher plant materials (Knights, 1968; Kvanta, 1968) that in some cases may be anthropogenically introduced (Aceves et al., 1991). In the winter samples the sterols predominate over the linear alcohols exhibiting a composition that ~is particle-size dependent, with cholest-5-en-3fl-ol and 24-ethylcholest-5-en-3fl-ol maximizing the sterol distribution in the coarser and the finer fractions, respectively. In addition to the sterols, several unsaturated sterones have been identified, namely cholesta-3,5-dien-7-one (No. 178) and 24-

Particle size screening of organic compounds

ethylcholest-4-en-3-one (No. 185). These ketones are oxidation products of formerly synthesized sterols (Holland and Diakow, 1979), their occurrence in the urban aerosol is likely related to contributions from soil dust (Grimalt and Saiz-Jimenez, 1989). The presence of triterpenoid alcohols, namely ~and fl-amyrin (Nos. 184 and 183, respectively) is also representative of terrestrial plant inputs (Simoneitet al., 1988). Acids. The fatty acids constitute another major group of solvent extractable compounds present in these aerosol fractions. They constitute distributions of even carbon numbered C ]4-C 1s saturated straight chain homologues occurring together with n-hexadecenoic (No. 143) and n-octadecenoic (No. 147) acids. They exhibit no major quantitative differences between the coarse and the fine particles being predominant compounds in the less concentrated extracts of the > 7.2/~m aerosol fractions. The presence of these fatty acid mixtures in atmospheric samples is generally attributed to inputs from microorganisms (Simoneit et al., 1988). The fatty acids identified in these samples are present in unbounded form, neither the samples neither the extracts have been submitted to hydrolysis in the experimental procedures. These free fatty acids exhibit distributions of homologues that do not parallel the acid moieties of the above described wax or noctyl esters, a disagreement that is consistent with the different origin of these compounds. Conversely, the saturated and unsaturated C16 and C]s n-alkanoic homologues, the major free fatty acids, show a similar composition as that of the monoacylglycerols. This

259

parallelism is in agreement with their presumed microorganism origin and suggest that the free fatty acids may have originated from hydrolysis of these monoglycerides. However, inputs related with decomposition of di- or triglycerides cannot be excluded because the analytical methods used in this study were not aimed to their identification. To the best of our knowledge this is the first time that the occurrence of monoacylglycerols in aerosol samples is reported. The

Table 3. Principal ions in the mass spectra of the TMS derivatives of 1-octadecanoic and l-octadecenoic glycerols Ion structure Molecular group [M] + [M- 15] + [M-73] + [MOO] + [M- 103] + [M-88-73] + RCO group [RCO + 74] + [RCO] + TMS-glycerol group [(CH3)3SiOCH-CH2-OSi (CH3)3] + [(CH 3)3SIOC = CHOSi (CH3)3] + [CH2 =CH(OH)CH2OSi (CHa)3] + [CH 2 = C H C H =OSi (CH3)3] + [CH2OSi (CHa)a] + [Si (CHa)a] +

18 : 1

18 : 0

500 485 427 410 397 339

502 487 429 --* 399 341

339 265

341 267

205

205

203

203

147 129 103 73

147 129 103 73

*Not detected.

T3

(a)

4O0OOO

129 ~ 3cx~

c O' "o

.~ 200 000 <~ I00 000

103

L 147

597

/

203

L ,lEt,

....

100

~. .L a~

360

~

~o

Moss/Charge

5OO0OO

c cI

399

73

(b)

4O0OOO 147

<~ 200000 I00 000

103 I ~

,L.JL,a I00

203 267 ~

341 . L.

"

429

487

L

400

Moss/Chorge

Fig. 4. Mass spectra of the trimethylsilyl ether derivatives of 1-n-octadecenoylglycerol (a) and 1-noctadecanoylglycerol (b) identified in the polar fractions of the Barcelona aerosols.

260

M. ACEVESand J. O. GR1MALT

mass spectra of the n-octadecanoyl and n-octadecenoyl glycerols are shown in Fig. 4 and the structural assignment of the principal ions is reported in Table 3. The intense M-103 ion (base peak in the saturated compound) and the 103/129 ratio lower than unit are characteristic of acyl substitution at position 1 (Myher

et al., 1974). N o 2-monoacylglycerols have been identified in these aerosols. In contrast with these distributions, the occurrence of other acids such as dehydroabietic acid (No. 154), and the corresponding methyl ester (No. 151), is representative of contributions from coniferous veg-

Table 4. Polar compounds identified in the particulate fractions of the aerosol samples collected in Barcelona* No.

135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160

161 162 163 164 165 166 167 168

Compound

9H-Fluoren-9-one n-Tetradecanoic acid§ Di-iso-butyl phthalate 1H-Phenalen-l-one n-Pentadecanoic acid n=Hexadecan- 1-o1§ Di-n-butyl phthalate 9,10-Anthracenedione n-Hexadecenoic acid n-Hexadecanoic acid

173

174 175 176 177 178 179 180 181 182 183 184 185

186

180 285 149 180 299 299 149 180 311 313

Winter aerosol Coarser Fines

2.8

3.0

7.5

17

Summer aerosol Coarse Fine

1.5

12

1.1

I0

2.5 3.6

3.8 7.1

0.36 0.58

1.4 5.1

1.6 2.3 1.0

3.8 4.7 8.0

0.21 0.20 0.25

0.94 1.1 2.0

5.1

4H-Cyclopenta(def)phenanthren-4one n-Octadecan- 1-ol n-Octadecenoic acid n-Octadecanoic acid Benzyl butyl phthalate n-Eicosan-l-ol Methyl dehydroabietate 7H-Benzo(c}fluoren-7-one Dioctyl hexanedioate Dehydrobietic acid 11H-Benzo(b)fiuoren- 11-one 11H-Benzo(a)fluoren- 1l-one

1H-Benz(de)anthracen-7-one 7H-Benz(de)anthracen-7-one Di-n-octyl phthalate n-Docosan-l-ol 1-n-Hexadecenoylglycerol§ 1-n-Hexadecanoylglycerol Di-iso-nonylphthalate n-Tetracosan-l-ol 1-n-Octadecenoylglycerol 1-n-Octadecanoylglycerol

llH-Benz(bc)aceanthrylen-ll-one 4H-Cyclopenta(deJ)triphenylen4-one

169 170 171 172

Diagnostic ions

4H-Cyclopenta(deJ)chrysen-4-one n-Pentacosan- 1-ol 6H-Benzo(cd)pyren-6-one n-Hexacosan- 1-ol n-Heptacosan- 1-ol Coprostanol§ n-Octacosan- 1-ol Colest-5-en-3fl-ol n-Nonacosan-l-ol Cholesta-3,5-dien-7-one 24-Methylcholest-5-en-3#-ol 24-Ethylcholesta-5,22-dien-3fl-ol n-Triacontan-l-ol 24-Ethylcholest-5-en-3#-ol fl-Amyrin§ ct-Amyrin 24-Ethylcholest-4-en-3-one n-Dotriacontan- 1-ol

204 327 339 341 149 355 239 230 370 239 230 230 230 230 149 383 369 371 149 412 397 399 254 254 254 426 254 440 454 370 468 129 482 174 129 484 496 129 218 218 124 524

1.1

12

0.2

5.0

25

4.0

1.0 0,29

4.0 9.5

30

<0.1 1.2

4.9 1.6

1,1

12

3.9

3.7

0,69 5.8

9.5 16

2.5 0.35

2.4 1.2

nd 0.20 0,57 2.0

4.0 11 5.0 19

nd 0.06 1.7 0.13

nd 0.20 1.1 1.2

0.4

9.8

0.49

0.42

* Concentrations in ng m-3. t > 7.2 pm. ~ < 0.5 #m. § Acids and alcohols analysed at trimethylsilyl ether derivatives.

Particle size screening of organic compounds

261

171

1000 000

(o)

158

9OOOOO 156 coo o0o

152

7~0 00o

60oooo 0

<~

nooooo 4 o o o0o

138

155

30OOO0

169

168 I

200 000

I0o o0o Itl

0

e a"

.4.

,Ik

-w

,

"

'~m

.

m

i

.

III

I

20 Time

(rain)

171 900000'

(b) 800(300'

700000'

600000 '

"~c 500 000 ~

158

400 000

300000

156 169

200000 168

I

, I00000 . . . . . . . . .

0

Jl,_.~.~

J

~

,

,iJ,

~]..J

l

IO

Time

(rain)

Fig. 5. Merged m/z 180, 204, 230 and 254 mass fragmentograms showing the distribution of the PAH ketones in the fine particle size aerosol samples collected in winter (a) and in summer (b). The numbered peaks refer to Table 3.

etation. These compounds can be transformed into retene by photochemical oxidation in the atmosphere (Simoneit and Mazurek, 1982), a hydrocarbon that has also been identified in the urban aerosols from Barcelona (No. 68 in Table 2). Polycyclic aromatic ketones and quinones. Trace amounts of oxygenated PAH, namely ketones and some quinones, have been found in the polar fractions. The compounds identified are listed in Table 4 and their composition is represented in Fig. 5 by means of the merged molecular ion mass fragmentograms. These compounds are formed by oxidation of methyl-

ene groups attached to aromatic rings, a reaction that is produced during fuel combustion so that they are present in vehicular traffic exhausts and in wood and coal combustion emissions (Choudhury, 1982; Ramdhal, 1983b). The ketones found in highest concentration as result of the combustion emissions are 9H-fluoren-9-one, 4H-cyclopenta(def)phenanthren-4one, 7H-benz(de)anthracen-7-one and 6H-benzo(cd)pyren-6-one (Ramdhal, 1983b). These ketones have been found in the urban aerosols considered in this study (Nos 135, 145, 158 and 171, respectively)and the two latter constitute the most abundant species of this

262

M. ACEVESand J. O. GR1MALT

group of polycyclic aromatic compounds in all the aerosols studied. The lower abundance of ketones Nos. 135 and 145 is likely due to their partitioning between the gas and particulate phase, an effect that can also be observed in higher molecular weight species as revealed from the comparison of the summer and winter polycyclic aromatic ketone profiles (Fig. 5). In fact, the whole mixtures of polycyclic aromatic ketones follow the same distribution trend as the PAH being preferentially associated to the submicron particle size fractions and exhibiting a strong decrease of the lower molecular weight species in the warm seasons. CONCLUSIONS Petrogenic hydrocarbons are the dominant products in the G C profiles of the solvent-extractable organic fractions from large and small airborne particles collected in urban areas. They are essentially represented by an U C M of hydrocarbons, n-alkanes with no odd/even carbon number predominance, 17ct(H),21fl(H)-hopanes, and parent and alkylated dibenzothiophenes, phenanthrenes and chrysenes/triphenylenes. Pyrolytic products, namely parent PAH eluting between fluorene and coronene and polycyclic aromatic ketones, constitute the second major quantitative group of molecules that can be recognized in these chromatographic traces. The third major group of anthropogenic compounds is related with the plasticizers being constituted by phthalates and related esters. The largest group of biogenic products corresponds to higher plant lipids and constitute distributions of straight chain and polycyclic molecules. The former include odd carbon numbered n-alkanes and n-alkan2-ones, and even carbon numbered n-alkan-l-ols, naldehydes, dimethyl alkadienoates and saturated and unsaturated wax esters. The second is constituted by sterols, namely 24-ethylcholent-5-en-3fl-ol, ~- and flamyrin, retene, dehydroabietic acid and its corresponding methyl ester. Other biogenic sources leaving a fingerprint in these chromatograms are related to microbial contributions. They can be recognized by the distributions of predominant fatty acids and 1-nalkanoyl and 1-n-alkenoyl glycerols. Many of these biogenic molecules exhibit predominant peaks in the large particle size G C traces which is a consequence of the combined effect of the preferential occurrence of the combustion-related compounds in the submicron fraction and the higher plant lipids in the coarse particles. This effect is particularly noteworthy in summer reflecting the higher productivity of the natural precursors during the warm seasons. The summer aerosols also exhibit a strong decrease in the peaks corresponding to the more volatile compounds (aliphatic and aromatic hydrocarbons with molecular weight lower than 352 and 226, respectively) due to the gas-particle partition shift produced by the higher ambient temperatures.

REFERENCES

Aceves M., Grimalt J. O., Albaiges J., Broto F., Comellas L. and Gassiot M. (1988) Analysis of hydrocarbons in aquatic sediments. II. Evaluation of common preparative procedures for petroleum and chlorinated hydrocarbons. J. Chromatogr. 436, 503-509. Aceves M., Grimalt J. O., Sunyer J., Anto J. M. and Reed Ch.E. (1991) Identification of soybean dust as an epidemic asthma agent in urban areas by molecular marker and RAST analysis of aerosols. J. Allergy clin. Immunol. 88, 124-134. Ajuntament de Barcelona (1985) AnMisi diagnostica de la contaminaci6 atmosf~rica de Barcelona (in Catalan). Ajuntament de Barcelona. Albaiges J. and Guardino X. (1980) Gas chromatographic-mass spectrometric identification of alkylcyclohexanes and cyclohexenes. Chromatographia 13, 755-762. Anto J. M., Sunyer J., Grimalt J. O., Aceves M. and Reed C. E. (1989a) Outbreak of asthma associated with soybean dust. N. Eng. J. Med. 321, 1128. Anto J. M., Sunyer J., Rodriguez-Roisin R., Suarez-Cervera M. and Vazquez L. (1989b) Community outbreaks of asthma associated with inhalation of soybean dust. N. Eng. J. Med. 320, 1097-1102. Back S. O., Field R. A., Goldstone M. E., Kirk P. W, Lester J. N. and Perry R. (1991) A review of atmospheric polycyclic aromatic hydrocarbons: sources, fate and behaviour. Water, Air, Soil Pollut. 60, 279-300. Bird C. W. and Lynch J. M. (1974) Formation of hydrocarbons by microorganisms. Chem. Soc. Rev. 3, 309-328. Choudhury D. R. (1982) Characterization of polycb,clic ketones and quinones in Diesel emission particulates by gas chromatography/mass spectrometry. Envir. Sci. Technol. 16, 102-106. Cranwell P. A. (1977) Organic geochemistry of Cam Loch (Shutherland) sediments. Chem. Geol. 20, 205-221. Eglinton G. and Hamilton R. (1967) Leaf epicuticular waxes. Science 156, 1322. Farrington J. W. and Meyers P. A. (1975) Hydrocarbons in the marine environment. In Environmental Chemistry (edited by Eglinton G.), Vol. 1, pp. 109-136. The Chemical Society, London. Furuhama S. and Hiruma M. (1979) Unusual phenomena in engine oil consumption. In Proceedings of ASLE/ASME Lubrication Conference (edited by Cooke M. and Dennis A. J.). ASLE, 1979. Greenberg A. (1989) Phenomenological study of benzo(a)pyrene and cyclopenteno(cd)pyrene decay in ambient air using winter/summer comparisons. Atmospheric Environment 23, 2797-2799. Greenberg A., Botzzelli J. W., Cannova F., Forstner E., Giorgio P., Stout D. and Yokoyama R. (1981) Correlations between lead and coronene concentrations at urban, suburban and industrial sites" in New Jersey. Envir. Sci. Technol. 15, 566-570. Grimalt J. O. and Albaiges J. (1990) Characterization of the depositional environments of the Ebro Delta (western Mediterranean) by the study of sedimentary lipid markers. Mar. Geol. 95, 207-224. Grimalt J. O. and Saiz-Jimenez C. (1989) Lipids of soil humic acids. 1. The hymatomelanic acid fraction. Sci. Total Envir. 81/82, 409-420. Grimmer G. and Hildebrandt A. (1975) Investigations on the carcinogenic burden by air pollution in man XIII. Assessment of the contributions of passenger cars to air pollution by carcinogenic polycyclic hydrocarbons Zbl. Bakt. Hyg. (I. Abt. Orig.) B161, 104-124. Grimmer G., Jacob J., Dettbarn G. and Naujack K.-W. (1985) Determination of polycyclic aromatic hydrocarbons, azaarenes and thiaarenes emitted from coal-fired residential furnaces by gas chromatography/mass spectrometry. Fresenius Z. Anal. Chem. 322, 595-602.

Particle size screening of organic compounds Grimmer G., Naujack K.-W. and Schneider D. (1980) Changes in PAH profiles in different areas of a city during the year. In Polynuclear Aromatic Hydrocarbons: Chemistry and Biological Effects (edited by Bjorseth A. and Dennis A. J.), pp. 107-125. Battelle Press, Columbus, OH. Grimmer G., Jacob J., Naujack K.-W. and Dettbarn G. (1983) Determination of polycyclic aromatic compounds emitted from brown-coal-fired residential stoves by gas chromatography/mass spectrometry. Analyt. Chem. 55, 892-900. Guicherit R. and Schulting F. L. (1985) The occurrence of organic chemicals in the atmosphere of Los Angeles. Sci. Total Environ. 43, 193-219. Hampton C. V., Pierson W. R., Schuetzle D. and Harvey T. M. (1983) Hydrocarbon gases emitted from vehicles on the road. 2. Determination of emission rates from Diesel and Spark-Ignition vehicles. Envir. Sci. Technol. 17, 699-708. Holland H. L. and Diakow P. R. P. (1979) Microbial hydroxylation of steroids. 5. Metabolism of androst-5-en3,17-dione and related compounds by Rhizopus arrhizus ATCC 11145. Canad. J. Chem. 57, 436-440. Kamens R. M., Guo Z., Fulcher J. N. and Bell D. A. (1988) Influence of humidity, sunlight, and temperature on the daytime decay on polyaromatic hydrocarbons on atmospheric soot particles, Envir. Sci. Technol. 22, 103-108. Katz M. and Chan C. (1980) Comparative distribution of eight polycyclic aromatic hydrocarbons in airborne particulates collected by conventional high-volume sampling and by size-fractionation. Envir. Sci. Technol. 14, 838-843. Knights B. A. (1968) Studies in the Cruciferae: sterols in pollen of Brassica napus. Phytochemistry 7, 1707-1708. Kolattukudy P. E. (1970) Plant waxes. Lipids 5, 259-275. Kvanta E. (1968) Sterols in pollen. Acta chem. Scand. 22, 2161-2165. LaFlamme R. E. and Hires R. A. (1978) The global distribution of polycyclic aromatic hydrocarbons in recent sediments. Geochim. cosmochim. Acta 42, 289-303. Miguel A. H. and Friedlander K. S. (1978) Distribution of benzo(a)pyrene and coronene with respect to particle size in Pasadena aerosol in the submicron range. Atmospheric Environment 12, 2407-2417. Myher J. J., Marai L. and Kuksis A. (1974) Identification of monoacyl- and monoalkylglycerols by gas-liquid chromatography-mass spectrometry using polar siloxane liquid phases. J. Lipid Res. 15, 586-592. National Research Council (1983) Polycyclic Aromatic Hydrocarbons: Evaluation of Sources and Effects. National Academic Press, Washington, DC. Nielsen T. (1988) The decay of benzo(a)pyrene and cyclopenteno(cd)pyrene in the atmosphere. Atmospheric Environment 22, 2249-2254. Pierce R. C. and Katz M. (1975) Dependency of polynuclear aromatic hydrocarbon content on size distribution of atmospheric aerosols. Envir. Sci. Technol. 9, 347-353. Pilinis C., Seinfeld J. H. and Seigneur C. (1987) Mathematical modeling of the dynamics of multicomponent atmospheric aerosols. Atmospheric Environment 21, 943-953. Pistikopoulos P., Wortham H. M., Gomes L., Masclet-Beyne S., Bon Nguyen E., Masclet P. A. and Mouvier G. (1990) Mechanisms of formation of particulate polycyclic aromatic hydrocarbons in relation to the particle size distribution: effects on meso-scale transport. Atmospheric Environment 24A, 2573-2584.

263

Prahl F. G. and Pinto L. A. (1987) A geochemical study of long-chain n-aldehydes in Washington coastal sediments. Geochim. cosmochim. Acta 51, 1573-1582. Ramdahl T. (1983a) Retene- a molecular marker of wood combustion in ambient air. Nature 306, 580-582. Ramdahl T. (1983b) Polycyclic aromatic ketones in environmental samples. Envir. Sci. Technol. 17, 666-670. Ratajczak E.-A., Ahland E., Grimmer G. and Dettbarn G. (1984) Verminderung der Emission Von Polycyclischen Aromatischen Kohlenwasserstoffen bein Ersatz von Pech durch Bitumen in Steinkohlenbriketts. Staub.-Reinhalt. Luft. 44, 505 509. Ritsema R., Cofino W. P., Frintrop P. C. M. and Brinkman U. A. Th. (1989) Trace-level analysis of phthalate esters in surface water and suspended particulate matter by means of capillary gas chromatography with electron-capture and mass-selective detection. Chemosphere 18, 2161-2175. Russel D. J. and McDuffie B. (1983) Analysis for phthalate esters in environmental samples. Separation from PCBs and pesticides using column liquid chromatography. Int. J. envir, analyt. Chem. 15, 165-183. Sicre M. A., Marty J. C., Saliot A., Aparicio X., Grimalt J. O. and Albaiges J. (1987) Aliphatic and aromatic hydrocarbons in different sized aerosols over the Mediterranean Sea: occurrence and origin. Atmospheric Environment 21, 2247-2259. Simoneit B. R. T. (1984) Organic matter of the troposphere IlL Characterization and sources of petroleum and pyrogenic residues in aerosols over the Western United States. Atmospheric Environment. 18, 51-67. Simoneit B. R. T. (1985) Application of molecular marker analysis to vehicular exhausts for source reconcilations. Int. J. envir, analyt. Chem. 29, 203 233. Simoneit B. R. T. and Mazurek M. (1982) Organic matter of the troposphere II. Natural background ofbiogenic lipid matter in aerosols over the rural western United States. Atmospheric Environment 16, 2139 2159. Simoneit B. R. T., Cox R. E. and Standley L. J. (1988) Organic matter of the troposphere-- IV. Lipids in Harmattan aerosols of Nigeria. Atmospheric Environment 22, 983-1004. Takada H., Onda T. and Ogura N. (1990) Determination of polycyclic aromatic hydrocarbons in urban street dusts and their source materials by capillary gas chromatography. Envir. Sci. Technol. 24, 1179-1186. Tancrede M., Wilson R., Zeise L. and Crouch E. A. C. (1987) The carcinogenic risk of some organic vapors indoors: a theoretical survey. Atmospheric Environment 21, 2187 2205. Trier C. J., Petch G. S., Rhead M. M., Fussey D. E. and Shore P. P. (1987) Assessing the contribution of lubricating oil to diesel exhaust particulates using geochemical markers. In Vehicle Emissions and their Impact on European Air Quality, Proceedings IMechE, C341/87, pp. 109-117. Trier C. J., Coates D. S., Fussey D. E. and Rhead M. M. (1990) The contribution of lubricating oil to diesel exhaust emissions. Sci. Total Envir. 93, 167 174. Van Vaeck L. and Van Cauwenberghe K. (1978) Cascade impactor measurements of the size distribution of the major classes of organic pollutants in atmospheric particulate matter. Atmospheric Environment 12, 2229-2239. Van Vaeck L. and Van Cauwenberghe A. (1985) Characteristic parameters of particle size distributions of primary organic contituents of ambient aerosols. Envir. Sci. Technol. 19, 707 716.