Jd000278 1.12

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 107, NO. D8, 10.1029/2000JD000278, 2002 Particle size distribution of organic primary and secondary aerosol constituents in urban, background marine, and forest atmosphere Ilias G. Kavouras and Euripides G. StephanouEnvironmental Chemical Processes Laboratory, Department of Chemistry, University of Crete, Heraklion, Hellas, Greece Received 19 December 2000; revised 20 August 2001; accepted 23 August 2001; published 26 April 2002.
Polynuclear aromatic hydrocarbons (PAHs), n-alkanes, n-alkanals, n-alkanols, saturated and unsaturated carboxylic acids, a, w-dicarboxylic acids, and carbonyl and carboxylic photooxidationproducts of monoteprenes were determined in particle-sized aerosols of urban (Heraclion, Island ofCrete, Greece), background marine (Island of Crete, Greece), and forest (Northern Greece andPortugal) atmospheres. The n-alkanes were mostly associated with fine particles in the urban andforest aerosol, and their mass mean aerodynamic diameter (MMAD) calculated over the whole sizerange (total MMAD) was 0.45 mm and 0.63 mm, respectively. In the background marine aerosol, n-alkanes were more evenly distributed, and their MMAD was 2.00 mm, because of physical changesoccurring during their long-range transport. Similar observations have been done for PAHs and n-alkanals. Conversely, the most biogenic compound class, namely n-alkanols, were evenlyassociated in the urban, background marine, and forest aerosol, between fine and coarse particles,and their corresponding total MMAD was 2.45, 2.69, and 1.67 mm, respectively. The total MMADof n-alkanoic acids was 0.71, 0.62, and 0.91 mm in the urban, background marine, and forestaerosol, respectively. Several compounds associated with photochemical reactions in theatmosphere were detected in urban marine and forests aerosol in the fine and ultrafine fraction,showing the low total MMAD (0.28 – 0.77 mm) in all aerosol types.
Atmospheric Composition and Structure: Constituent sources and sinks; 0365 AtmosphericComposition and Structure: Troposphere—composition and chemistry; 0305 AtmosphericComposition and Structure: Aerosols and particles (0345, 4801) [Kavouras et al., 1998b, 1999a, 1999b]. Information concerningthe physical and chemical changes of aerosol and its removal from [2] The organic fraction of particulate matter is an important and the atmosphere can be found in the literature [Van Vaeck and Van complex matrix of aerosol. It frequently accounts for more than 30% Cauwenberghe, 1985; Sicre et al., 1990a; Aboulkassim and Simo- of aerosol mass over continental populated and remote areas neit, 1996; Fernandes et al., 1999].
[Andreae et al., 1988; Talbot et al., 1992; Chow et al., 1993, [3] The study of the size distribution of both primary and 1994]. The total particulate organic carbon, from direct emissions secondary organic constituents of aerosol is an important factor or formed as secondary organic aerosol from atmospheric reactions to assess their role in climate forcing [Turpin et al., 2000], the of volatile organic compounds, has been estimated at $1 – 5 Â 1012 health hazards involved with particle inhalation [Pagano et al., g [Jaenicke, 1978; Hahn, 1980]. Although organic compounds 1996], and the deposition potential of atmospheric organic carbon.
constitute a large amount of the total dry atmospheric fine-particle Particle size distributions of n-alkanes and polynuclear aromatic mass, their concentration and formation mechanisms are less hydrocarbons (PAHs) have been mostly studied in the urban understood than those of sulfate and nitrate [Turpin et al., 2000].
environment [Van Vaeck and Van Cauwenberghe, 1985; Aceves The ability of organic aerosols to act as cloud condensation nuclei and Grimalt, 1993; Allen et al., 1997; Kavouras et al., 1998a], and (CCN) seems to be closely related to the chemical properties of in some cases in the marine environment [Sicre et al., 1987, their constituents (e.g., polarity, solubility, etc.) and to the physical 1990b]. Studies on the size distribution of polar compounds properties of their particles (e.g., size, surface tension, etc. [Turpin (mainly n-alkanoic acids) are very scarce [Van Vaeck and Van et al., 2000]). A large number of studies focusing on the chemical Cauwenberghe, 1985]. The importance of polar organic aerosol characterization of the aerosol organic fraction, emitted from constituents has been recognized because of their role in atmos- different sources, were conducted [Stephanou, 1989, 1992; Stepha- pheric phenomena such as homogeneous nucleation [Tao and nou and Stratigakis, 1993; Rogge et al., 1993a, 1993b, 1993c, McMurry, 1989] and impaction of heterogeneous chemistry [Cruz 1993d, 1994, 1997a, 1997b, 1998; Lowenthal et al., 1994; Gogou et and Pandis, 1998]. The effect of polar organic compounds on the al., 1996; Pagano et al., 1996; Kavouras et al., 1998a, 1999a; formation of clouds was also recently investigated [Ansari and Simoneit, 1999]. The mechanisms of the chemical reactions Pandis, 2000; Cruz and Pandis, 1998; Russell et al., 1994].
between a variety of unsaturated hydrocarbons and oxidants, and [4] In a previous paper we reported a detailed study of the particle thus their ability to form secondary polar organic aerosol, were size distribution of urban indoor and outdoor aerosol anthropogenic investigated in detail [Atkinson, 1990; Atkinson and Arey, 1998; organic compounds such as n-alkanes, branched alkanes (environ- Pandis et al., 1992]. Furthermore, the chemical coupling of mental tobacco smoke molecular markers), and PAHs [Kavouras et specific polar organic compounds with the formation of new al., 1998a]. Here we report a detailed study on the size distribution of particles formed over forests was established recently in situ primary and secondary organic constituents of urban, backgroundmarine, and forest aerosol collected in different locations of southern Copyright 2002 by the American Geophysical Union.
Europe. The major goal of the present research was to study the particle size distribution and the atmospheric deposition potential for KAVOURAS AND STEPHANOU: SIZE DISTRIBUTION OF ORGANIC AEROSOL Table 1. Total Suspended Particle (TSP), Extractable Organic Matter (EOM), and Molecular Markers Concentration Rangesa a Other abbreviations are as follows: UCM, unresolved complex mixture concentrations; NA, n-alkane concentration; ND, not detected.
a variety of molecular markers characterizing both primarily emitted described by Kavouras et al. [1998b, 1999a, 1999b]. At the above and secondarily formed organic aerosols. A series of primary mentioned sampling sites, aerosol samples were also collected, organic aerosol constituents such as aliphatic hydrocarbons, PAHs, with conventional high-volume samplers, in order to determine the n-alkanals, n-alkanols, n-alcanoic acids, and unsaturated carboxylic organic aerosol constituents concentration ranges (Table 1) [Gogou were identified and studied. Secondary organic aerosol constituents et al., 1996; Kavouras et al., 1998a, 1998b, 1999a, 1999b].
containing dicarboxylic acids and carbonyl and carboxylic com-pounds produced through the photooxidation of volatile biogenic Fractionation, Derivatization, and Identification hydrocarbons were also studied. Taking into consideration the [7] A detailed description of the analytical procedure used for limited particle size resolution of high-volume cascade impactors, extraction, separation, and analysis of the main lipid fractions has we subjected our results to a phenomenological discussion of the been published elsewhere [Gogou et al., 1998]. Briefly, filters were aerosol generation processes, as well as of their physical and extracted by refluxing methylene chloride for 20 hours. Each organic extract was evaporated using the Kuderna-Danish method,transferred to a 1-mL vial, dried under a gentle stream of nitrogen, and weighted to obtain for each filter the extractable organic matter(EOM). The EOM was then dissolved in a small aliquot of n- hexane and applied on the top of a glass column containing [5] A five-stage (plus backup filter) Sierra Andersen Model 230 specially treated silica gel. Nitrogen pressure was adapted to elute Impactor was used mounted on a high-volume pump (GMWL- the different compound classes: (1) n-hexane for aliphatics, (2) 2000, General Metal Works, Ohio, USA). Aerosol particles were toluene/n-hexane for PAHs, (3) n-hexane/methylene chloride for separated into six size fractions on glass-fiber filters, according to carbonyl compounds, (4) ethyl acetate/n-hexane for hydroxyl their aerodynamic cutoff diameters at 50% efficiency. Namely, the compounds, and (5) a solution of pure formic acid in methanol first stage is >7.2 mm, the second stage is 7.2 – 3.0 mm, the third for carboxylic acids. The individual fractions were spiked with stage is 3.0 – 1.5 mm, the fourth stage is 1.5 – 0.96 mm, the fifth internal standards (1-chlorohexadecane for n-alkanes, n-alkanals, stage is 0.96 – 0.5 mm, and the backup filter is <0.5 mm at 1.13 m3 n-alkanol silyl ethers, and for fatty acid methyl esters; hexame- minÀ1 flow rate. After collection, filters were placed in glass tubes thylbenzene for PAH and n-hexacosane and long-chain alkenones) and stored in a freezer at À30°C until extraction and analysis.
for quantitative determinations. Relative response factors, in both [6] Six urban size-distributed samples were collected, with a gas chromatography-mass spectrometry (GC-MS) and GC-MS in high-volume sampler, in the urban area of the city of Heraclion the selected ion monitoring mode, were calculated for 3 – 10 (150,000 inhabitants) [Kavouras et al., 1998a] during the fall of standard compounds, representing each compound class, of 1995. Six samples from a background marine site were collected at increasing molecular mass. Relative response factors for PAHs Environmental Chemical Processes Laboratory sampling station were calculated for each single compound individually. Control of (Finokalia, on the northern coast of Crete, 25°600E, 35°240N) from procedural blanks has been performed to assess possible contam- November 1996 to June 1997. The nearest largest city is Heraclion, ination. The total blank weight never exceeded 2% of the individ- located 70 km westward of Finokalia. The station is located at the ual sample extracts (except for the n-alkanoic acids, where the top of a hilly elevation (130 m) facing the sea within the sector of maximum contamination represented 10% of the total fraction 270° to 90° [Gogou et al., 1996; Mihalopoulos et al., 1997].
extract, especially for the homologues C14, C16, and C18). The Anthropogenic activities are negligible at a distance shorter than 20 contaminants were characterized by GC-MS analysis and compar- km within the above-mentioned sector. Samples were collected for ison with standard mixtures. The most frequent contaminants were 24 hours under north and south winds which are related with transport from continental Europe and Sahara, respectively [Miha- [8] The efficiency of the whole procedure (extraction, fractio- lopoulos et al., 1997]. Nine 12-hour (from 6:00 a.m. to 6:00 p.m.
nation, and derivatization) was very satisfactory. We obtained and from 6:00 p.m. to 6:00 a.m.) size-distributed aerosol samples recoveries (threefold measurements and standard deviation of less were collected in two forests: (1) an Eucalyptus forest located in than ±3%) of 61.5% up to 98.2% for n-alkanes, 60.3% up to 100% Ta´bua (Portugal) situated 100 km inland from the Atlantic coast for PAHs, 75% for n-alkanals, and 88.2% up to 97.7% for n- [Kavouras et al., 1998b, 1999a] from July to August 1996 and (2) a conifer forest (mainly Abies Borissi-Regis) located in Pertouli [9] All samples were analyzed on a Finnigan GCQ ion trap gas (1300 m) on the Agrafa mountains in central Greece [Kavouras et chromatography-mass spectrometry in the electron impact and al., 1999b] from July to August 1997. The secondary organic occasionally methane-chemical ionization mode [Kavouras et al., aerosol formation versus primary organic aerosol emission and the 1999a] equipped with a HP-5MS capillary column (30 m  0.25 parameters (e.g., ozone, OH radical, reactive nonmethane hydro- mm ID  0.25 mm film thickness). The temperature ramp for a carbons, etc.) influencing these processes were thoroughly splitless injection was at 270°C. The chromatographic column KAVOURAS AND STEPHANOU: SIZE DISTRIBUTION OF ORGANIC AEROSOL Table 2. Concentration Distribution and Mass Median Aerodynamic Diameter of Particulate Mass and Extractable Organic Mattera a Particle size is defined by Sierra Andersen cascade impactor. Mass median aerodynamic diameters are calculated for the whole range of impactor particle sizes (total) and for fine and coarse particles.
oven temperature was held at 70°C for 2 min. Then, the temper- other half is larger) was stepwise calculated using the following ature was increased to 150°C at a rate of 10°C minÀ1, and then to 290°C at a rate of 5°C minÀ1. At the end the oven was held at290°C for 30 min. Mass spectrometer ion source temperature was 200°C, and electron impact ionization potential was 70 eV. The identification was performed by interpretation of mass spectra and by using reference standards for mass spectra comparison (Dr.
(micrometers) and D1 is the lower particle size (micrometers) for 1. The odd carbon preference indices (CPI) for n-alkanes was i-impactor stage; Ci, Cj, and Ctotal are the mass concentrations (ng calculated as follows [Gogou et al., 1996]: mÀ3) for i- and j-impactor stages and total mass concentration,respectively. If DMMAD was higher than the upper particle sizecollected by the i-impactor stage, the calculation was repeated for The n-Alkanes originating from epicuticular wax of terrestrial plantsexhibit high values of CPI (CPI > 1). Conversely, CPI values for n-alkanes originating from vehicular emissions and other anthropo- genic activities are close to unity (CPI % 1).
2. The wax n-alkanes concentration (WNA) was calculated for [10] Concentration ranges of total suspended particles (TSP) each n-alkane as follows [Gogou et al., 1996]: and EOM of the aerosol collected in the urban, background marine,and forest sampling sites are presented in Table 1. In Table 1 the concentration ranges of the organic aerosol constituents, such as n- alkanes, PAHs, n-alkanals, n-alkanols, n-alkanoic acids, n-alkenoicacids, dicarboxylic acids, and the concentration ratio of unresolved Negative values of Cn were taken as zero. The percentage of total complex mixture of branched hydrocarbons (UCM) and n-alkanes wax n-alkanes to total n-alkanes (%WNA) was calculated as investigated in this study are also reported. For the forest aerosol the concentration ranges of pinonaldehyde, nopinone, and pinonicacid, characteristic photo-oxidation products of monoterpenes, are [11] Elevated total suspended particulate (TSP) concentration levels (100.0 – 192.0 mg mÀ3; Table 1) were determined in the n is the total concentration of wax n-alkanes and ÆNA is the total concentration of n-alkanes.
urban environment. The TSP concentration range in Heraclion was 3. The size distribution of particles (n0 similar to that reported in Barcelona [Aceves and Grimalt, 1993].
The TSP concentration range in the background marine site (14.0 – diagrams as follows [Van Vaeck and Van Cauwenberghe, 37.0 mg mÀ3; Table 1) and the forest sites (1.5 – 65.0 mg mÀ3; Table 1) were significantly lower than the corresponding ones of the urbansite. TSP concentration in the forest sites exhibited the highest variability (Table 1). This higher variability was also observed for the EOM of the background marine aerosol (0.5 – 8.0 mg mÀ3; Table 1) and the forest aerosol (0.1 – 22.0 mg mÀ3; Table 1) and toa lesser extent for the urban aerosol (12.0 – 34.0 mg mÀ3; Table 1).
where C is the concentration (ng mÀ3) for a given stage, Da is The impactor stage fractionation (Table 2) shows that urban aerosol the aerodynamic diameter (micrometers), and CT is the total can be described by two size-dependent modes. The first concentration (ng mÀ3) of a compound.
mode (fine particles <2.5 mm) corresponds to a mass median 4. Since cascade impactors classify and collect particles in size aerodynamic diameter (Table 2) of 0.67 mm. This value is higher ranges, mass mean aerodynamic diameter (MMAD) (particle than the corresponding one determined in the Barcelona aerosol diameter where one half of the particle mass is smaller and the (0.36 – 0.40 mm [Aceves and Grimalt, 1993]). The second mode KAVOURAS AND STEPHANOU: SIZE DISTRIBUTION OF ORGANIC AEROSOL Table 3. Organic Molecular Markers Concentration and Diagnostic Parameters Determined in Each Impactor Stage for Urban Aerosola a Abbreviations are as follows: Cn-Cm, homologue concentration ranges; Cn,max, homologues with the maximum concentration for molecular markers; CPI, carbon preference index; UCM, unresolved complex mixture; NA, total n-alkanes; WNA, leaf wax n-alkanes; PAHs, polynuclear aromatichydrocarbons; CPAHs, combustion-derived PAHs; TPAHs, total PAHs concentration; TPhs, total phenanthrenes.
(coarse particles >2.5 mm), for the urban (city of Heraclion) MMAD of 5.36 mm for PM and 5.13 mm for EOM. The smaller aerosol, corresponds to a MMAD of 5.96 mm. This value is in particle stages (<1.5 mm) in the forest aerosol contained most of the the same order of magnitude as that determined in Barcelona (4.6 – PM ($76%, calculated from Table 2) and EOM (81%, calculated 6.0 mm). The smaller particle stages (<1.5 mm) in urban aerosol from Table 2) mass. The concentrations of particulate matter contain most of the TSP mass ($71%, calculated from Table 2).
(4.56 – 7.81 mg mÀ3; Table 2) and extractable organic carbon The EOM was also described by two size-dependent modes in the (0.74 – 1.48 mg mÀ3; Table 2) associated with large particles urban aerosol. EOM was distributed between fine particles with a (Da > 1.5) were lower than those measured for smaller particles MMAD of 0.60 mm (Table 2) and coarse particles with a MMAD (Da < 1.5) (particulate matter is 6.56 – 38.53 mg mÀ3 and organic of 5.59 mm (Table 2). In the urban aerosol the smaller particle carbon is 1.75 – 13.07 mg mÀ3; Table 2). These results suggest that stages (<1.5 mm) contained most of the EOM mass ($80% the increase of aerosol mass is due to the higher contribution of calculated from Table 2). MMDA calculated for the whole range organic compounds. These compounds can be associated with of impactor sizes is 0.81 mm (Table 2) for particulate matter (PM) either direct emissions of lipids from the trees or condensation of and 0.69 mm (Table 2) for EOM. Forest aerosol differed from the low-vapor organic gases formed through the photooxidation of urban one when we consider the above parameters. MMDA unsaturated hydrocarbons or both [Kavouras et al., 1998b, 1999a, calculated for total particle size range for forest aerosol was 0.48 1990b]. The MMAD of an organic compound is found at a mm (Table 2) for PM and 0.41 mm (Table 2) for EOM. For forest significantly smaller particle size than for the total aerosol if a aerosol the impactor stage fractionation showed that fine particles condensation mechanism prevails. This should be the case in the (<2.5 mm) correspond to a MMAD of 0.39 mm for PM and 0.36 mm forest aerosol since the gas-particle conversion enriches the aerosol for EOM (Table 2). These values are considerably lower than fraction in the smaller particle size range [Kavouras et al., 1999b].
the corresponding ones determined in urban aerosol (Table 2).
[12] Organic aerosol molecular markers such as aliphatic hydro- Coarse particles (>2.5 mm) for forest aerosol correspond to a carbons, unresolved complex mixture (UCM) of branched hydro- KAVOURAS AND STEPHANOU: SIZE DISTRIBUTION OF ORGANIC AEROSOL Table 4. Organic Molecular Marker Concentration and Diagnostic Parameters Determined in Each Impactor Stage for MarineBackground Aerosola carbons, PAHs, n-alkanals, n-alkanols, n-alkanoic, and n-alkenoic background marine samples the n-alkanal homologues distribution and dicarboxylic acids were identified in the samples of the studied was very similar to the corresponding pattern of n-alkanols. The sampling sites (Table 1). The influence of air masses origin similarity of the homologue distribution between n-alkanols and [Mihalopoulos et al., 1997] on TSP, EOM, and the above molec- n-alkanals indicated a close relationship and thus a common origin ular markers has been reported elsewhere [Gogou et al., 1996]. The (terrestrial higher plant wax) of these two compound classes n-Alkanes dominated the aliphatic aerosol fraction of all samples.
[Gogou et al., 1996]. The concentrations of n-alkanals in the urban The study of these parameters confirmed a stronger input of area (5.5 – 7.0 ng mÀ3; Table 1) were higher than those measured in biogenic n-alkanes in background marine and forest aerosol than the background marine site (1.0 – 4.0 ng mÀ3; Table 1) and the forest in the urban samples. Lower UCM/n-alkanes concentration ratios areas (0.3 – 2.2 ng mÀ3; Table 1). The most abundant compound were observed (UCM/NA, Table 1) in the marine (0.0 – 6.0; Table 1) class determined among the neutral oxygenated lipids was that of n- and forest (2.2 – 7.0; Table 1) samples than in the urban ones (8.6 – alkanols (17.0 – 32.0 ng mÀ3 for urban, 3.0 – 17.0 ng mÀ3 for 14.0; Table 1), thus indicating a lower contribution of petroleum marine, and 0.1 – 40.0 ng mÀ3 for forest aerosol; Table 1). Carbox- hydrocarbons. In the forest samples the relative homologue distri- ylic acids were by far the most abundant compound class detected bution was correlated with this obtained from forest trees epicu- in all collected samples (Table 1). We detected four different groups ticular wax extract [Kavouras et al., 1999a]. The n-Alkanals, n- of carboxylic acids, namely n-alkanoic, n-alkenoic, dicarboxylic alkanols, and carboxylic acids were also identified in all samples acids [Stephanou and Stratigakis, 1993; Gogou et al., 1996], and a (Table 1). In the forest aerosol these lipids originated mainly from series of carboxylic acids considered as photooxidation products of epicuticular wax of forest trees [Kavouras et al., 1999b]. In urban monoterpenes [Kavouras et al., 1998a, 1999a, 1999b]. The latter and background marine samples the above lipids had a rather mixed were detected only in the forest atmosphere (0.2 – 71.0 ng mÀ3). The origin [Gogou et al., 1996]. In urban samples, n-alkanals homo- n-Alkanoic acids varied from 110.0 to 200.0 ng mÀ3 in the urban, logues distribution demonstrated a mixed origin from microbial 1.0 to 20.0 ng mÀ3 in the marine, and 14.0 to 194.0 ng mÀ3 in the sources and n-alkanes oxidation [Stephanou, 1989]. Conversely, in forest atmosphere (Table 1). The n-Alkenoic acids, because of their KAVOURAS AND STEPHANOU: SIZE DISTRIBUTION OF ORGANIC AEROSOL Table 5. Organic Molecular Marker Concentration and Diagnostic Parameters Determined in Each Impactor Stage for Forest Aerosola a For abbreviations, see Table 3; NC is not calculated.
higher reactivity [Stephanou and Stratigakis, 1993], were deter- Table 3) with particles <1.5 mm. Conversely, WNA showed a mined in lower concentration in the urban and marine atmosphere preference to particles >1.5 mm (Table 3). In the background (Table 1). In the forest atmosphere, because of their continuous marine aerosol a clear trend for the n-alkane size distribution was emission [Kavouras et al., 1999a], the concentration was higher not determined. The n-alkane concentration was more evenly (up to 103 ng mÀ3; Table 1). The organic aerosol molecular distributed between particle sizes (Table 4). The same observation markers described above were also identified in each impactor can be made for CPI, UCM, and %WNA (Table 4). These facts stage of the samples collected at the sampling sites (Tables 3 – 5).
probably reflect the physical changes (e.g., particle size changes [13] The analysis of the n-alkane homologues allowed the and wet and dry removal) of aerosol during long-range transport determination for each impactor stage of their relative distribution to the background marine sampling area [Gogou et al., 1996]. In (Cm À Cn), homologue with the maximum concentration (Cn, max), the forest aerosol, n-alkanes, rather, were associated ($82%, the UCM concentration, the CPI, and the percentage of leaf wax n- calculated from Table 5) with particles <1.5 mm. The CPI for all particle sizes was >2.0 (Table 5) and is reflected the predom- [14] In the urban aerosol the bulk ($81%, calculated from inant biogenic origin of the forest aerosol. UCM was associated Table 3) of n-alkane concentration was determined in particles mostly with particles <0.5 mm, while WNA were, rather, asso- with diameters <1.5 mm. The CPI of n-alkanes in all impactor ciated with particles >1.5 mm (Table 5). In order to reliably stages (Table 3) exhibited a mixed origin (petroleum residues and differentiate (between the three sampling sites) the n-alkanes higher plant wax) of n-alkanes. The n-alkanes associated with particle size distribution, Lundgren diagrams [Van Vaeck and Van particles <0.5 mm demonstrated the lowest CPI values (1.2 ± 0.3; Cauwenberghe, 1985; Aceves and Grimalt, 1993] were constructed Table 3). UCM was mostly associated ($86%, calculated from (Figure 1), and mass median aerodynamic diameters (MMAD) were KAVOURAS AND STEPHANOU: SIZE DISTRIBUTION OF ORGANIC AEROSOL Lundgren diagrams for (a) urban, (b) background marine, and (c) forest n-alkanes, wax n-alkanes concentration, and calculated for total particle size range and also separately for fine (<2.5 mm) and coarse (>2.5 mm) particles (Table 6). For the coarseparticles, n-alkanes MMAD did not differ greatly for the threeaerosol types (Table 6 [see also Aceves and Grimalt, 1993]. Con- versely, n-alkanes MMAD, for the total particle size range, was 0.45 mm for the urban aerosol and 2.00 mm and 0.63 mm for the back- ground marine and forest aerosols, respectively (Table 6). The corresponding values for the fine particles were 0.38 mm for the urban site, 0.81 mm for the background marine, and 0.49 mm for theforest sites (Table 6). The above data indicate a clear differentiation between the background marine aerosol and the urban and forestaerosols. This difference in the MMAD (higher values for marine aerosol) can be interpreted by the particle-size changes of aerosol during long-range transport in the background marine area [seealso Van Vaeck and Van Cauwenberghe, 1985]. The same effect can be observed when n-alkanes Lundgren diagrams are consid- ered (Figure 1). The maxima of the distribution are shifted toward larger particle sizes for both total n-alkanes and wax n-alkanes in the marine background aerosol (Figure 1b) in comparison with KAVOURAS AND STEPHANOU: SIZE DISTRIBUTION OF ORGANIC AEROSOL [15] Polynuclear aromatic hydrocarbons (PAHs) in the three different aerosol types exhibited a predominant occurrence in the particle size fraction <1.5 mm (93% for urban, 74% for marine, and 79% for the forest calculated from Tables 3, 4, and 5, respectively).
In the urban aerosol, PAHs had a MMAD for the total particle size range of 0.36 mm (Table 6) and predominance in the fraction of particles <0.5 mm ($70% as calculated from Table 3; Figure 2a).
This observation indicates that PAHs size distribution basically reflects the gas-to-particle condensation in the submicron rangeafter their emission to the urban atmosphere and subsequent cooling. This trend is more obvious if we consider the ratio of concentration of combustion-related PAHs (nine major nonalky-lated compounds: fluoranthene, pyrene, benz[a]anthracene, chrys- ene, benzofluoranthenes, benzo[a]pyrene, benzo[e]pyrene, indeno[1,2,3-cd]pyrene, and benzo[ghi]perylene) (CPAHs) to the corresponding concentration of total PAHs (TPAHs) (CPAHs/ TPAHs). The sum of parent PAHs expressed as CPAHs is used hereafter to represent pyrolytic PAH sources. The highest ratios are observed for particles <1.5 mm (0.67 – 0.82; Table 3). Conversely, the corresponding ratio of petrogenic PAHs (phenanthrene, methyl- and dimethyl-phenanthrene) (TPh) (TPh/TPAHs [Gogou et al., 2000]) to the total PAHs reached their highest values in the urban atmosphere in particles >1.5 mm (0.14 – 0.34; Table 3). The size distributions were consistent with the condensation of large non-volatile PAHs (e.g., compounds belonging to the CPAHs such as benzo[a]pyrene, benzo[e]pyrene, indeno[cd]pyrene, and benzo[g- hi]perylene) (Figure 2a) on small particles during cooling of exhaust in the urban environment. The smaller, semivolatile PAHs(e.g., phenanthrene, methyl- and dimethyl-phenanthrene) expressed as TPh here represent the petrogenic PAH sources (unburned fossil fuels) in this study [Gogou et al., 2000]. TPh became distributed between the smaller and larger particles (TPh/ TPAHs in Tables 3 and 5 and Figures 1a and 1c), probably via continuing vaporization and condensation processes in the urban and forest atmospheres. Venkataraman and Friedlander [1994] made similar observations for polynuclear aromatic hydrocarbons in ambient aerosols. The MMAD calculated for PAHs associatedwith fine particles in the urban site was 0.34 mm (Table 6). This value was very similar to the corresponding one calculated for the total particles size range (0.36 mm; Table 6). The PAHs MMAD for the total particle size range was 0.73 mm for the marine aerosol and 0.63 mm for the forest aerosol (Table 6). The effect of particle sizechanges, due to transport, is reflected on both marine [Gogou et Lundgren diagrams for (a) urban, (b) background al., 1996] and forest [Kavouras et al., 1998b] aerosol for PAHs.
marine, and (c) forest total polynuclear aromatic hydrocarbons Van Vaeck and Van Cauwenberghe [1985] have had similar (TPAHs), combustion-derived PAHs (CPAHs), and total phenan- [16] The n-Alkanals were detected in all impactor stages for the urban and the background marine aerosol. No detectable urban (Figure 1a) and forest aerosols (Figure 1c). Considering amounts of these compounds could be detected in the impactor Lundgren diagrams for individual n-alkanes with different sub- stages for forest aerosol (Tables 3 – 5). Considerable differences in aerosol composition and particle size distribution among n-alka- À 4.76 [Pankow, 1997]), n-nonadecane (log PoL À 6.87 [Pankow, nals were observed between urban and background marine areas.
1997]), and n-hentriacontane, we also observed the same phe- The n-Alkanals from C9 to C23, were detected in the urban site, nomenon as for total and wax n-alkanes: The maxima of the maximizing at C17 and C22 (Table 3). Higher molecular weight distribution were shifted toward larger particle sizes for the above n-alkanals from C18 to C28, maximizing at C26 and C28, were homologues in the marine background aerosol. In urban aerosol a identified in background marine aerosol (Table 4). The occur- difference in particle size distribution was observed between total rence of low molecular weight n-alkanals in the urban atmos- n-alkanes and wax n-alkanes (Figure 1a). The distribution max- phere can be explained by photooxidation reactions of imum for the wax n-alkanes corresponds to larger particle sizes anthropogenic hydrocarbons (such as alkenes) in urban areas (>1 mm) than the corresponding for total n-alkanes (<1 mm) [Atkinson, 1990]. Conversely, biogenic sources are on the origin (Figure 1a). This observation is concomitant with the measured of n-alkanals in the background marine atmosphere (Table 4 [see CPI values, which indicated a mixed origin for urban n-alkanes also Stephanou, 1989]). This is further supported by the strong associated with particles >1 mm and a more petroliferous origin even-to-odd predominance (CPI 2.0 – 8.1; Table 4). In urban for the submicron particles (Figure 1a). In the marine aerosol the aerosol, n-alkanals were primarily associated with ultrafine and Lundgren diagrams (Figure 1b) demonstrate similar particle size fine particles (<0.96 mm; Table 3). The concentration levels of distribution patterns for total n-alkanes and wax n-alkanes, as n-alkanals in all impactor stages were comparable (Table 3). The results of the mixing occurred during long-range transport to this Lundgren diagram (Figure 3a) of n-alkanals, suggests a lognor- mal distribution as a function of particle size. The MMAD KAVOURAS AND STEPHANOU: SIZE DISTRIBUTION OF ORGANIC AEROSOL the vicinity of the emission source to the sampling site. The calculated MMAD for n-alkanols associated to fine particles was 0.70 mm in the urban site, 1.05 mm in the background marine site,and 0.96 mm in the forest site (Table 6). The comparison ofMMAD of n-alkanols calculated for fine particles (Table 6) with the corresponding MMAD of n-alkanes, n-alkanals, and PAHsdemonstrate the predominant biogenic character of n-alkanols in relation to the other compound classes. The n-alkanols particle sizedistribution patterns were further studied by using Lundgren diagrams (Figure 3b). A clear bimodal particle size distribution pattern was observed for forest n-alkanols (Figure 3b), while in the urban and marine aerosols, n-alkanols, rather, were associated with [18] Minor differences in the composition of saturated acids were observed, as a function of particle size, among the three different aerosol types (Tables 3 – 5). The n-alcanoic acids ranged 8 to C30 (Tables 3 – 5). Palmitic acid (C16) was the predom- inant compound in the carboxylic fraction of all analyzed samples.
Low molecular weight acids such as octanoic (C8) and nonanoic (C9) acid were identified in the fine and ultrafine fractions (<1.5 mm) Lundgren diagrams for urban, background marine and forest (a) n-alkanals and (b) n-alkanols concentration.
calculated for n-alkanals for the total particle size range in the urban site was 0.44 mm (Table 6). The corresponding value forthe marine aerosol MMAD was 1.15 mm (Table 6). The changes for marine aerosol due to transport were also reflected on n-alkanalsMMAD. Thus the MMAD calculated for n-alkanals associated with fine particles in the urban site was 0.40 mm (Table 6). As for PAHs,this value was very similar to the corresponding value calculated for the total particles size range (0.44 mm; Table 6). The Lundgren size distribution pattern and the data of MMAD calculation indicate a strong effect of coagulation and agglomeration processes during n-alkanals residence in the urban atmosphere. The origin of n-alkanals in the marine aerosol has a more biogenic character, and this was reflected in their CPI (up to 6.2; Table 4) and high MMAD (1.00 mm for the fine particles; Table 6) values.
[17] The n-Alkanols were detected in all aerosol samples (Table 1). This compound class has mostly a biogenic origin, and therefore its composition and concentration levels in the three different aerosols did not differ greatly. The n-Alkanols rangingfrom C14 to C32, with maxima at C26 and C28, were detected inall particle sizes sampled in the urban area (Table 3). Their composition in background marine and forest aerosol samples was similar (C12 – C20 with maxima at C24 and C28 for the marine aerosol and C12 – C28 with maximum at C26 for the forest aerosol; Tables 4 and 5) to the composition of the urban ones. The high CPI values for all size fractions in urban (CPI 3.1 – 6.5; Table 3), background marine (CPI 2.9 – 11.1; Table 4), and forest aerosol (CPI >30.9; Table 5) strongly suggest direct emission from epicuticular waxes of higher terrestrial plants. The MMAD calculated for n-alkanols for the total particle size range in the urban site was 2.30 mm (Table 6). This value was the highest calculated among those for the other compound categories (Table 6). The corresponding MMAD for marine aerosol was 2.69 mm and 1.67 mm for the forest aerosol (Table 6). The effectof physical changes due to transport was reflected in a lower degree Lundgren diagrams for urban, background marine, and on marine aerosol for n-alkanols than for n-alkanals and n-alkanes forest (a) n-alkanoic, (b) n-alkenoic, and (c) dicarboxylic acid (Table 6). Forest n-alkanols had a lower MMAD probably due to KAVOURAS AND STEPHANOU: SIZE DISTRIBUTION OF ORGANIC AEROSOL Lundgren diagrams for forest (a) carboxyl and (b) carbonyl photooxidation products of monoterpenes of urban and background marine samples (Tables 3 and 4). Con- (from 0.3 to 1.0 ng mÀ3; Table 4) were lower than those measured versely, higher molecular weight carboxylic acids (>C24) were in urban (0.2 – 6.2 ng mÀ3; Table 3) and forest (from 3.5 to 9.0 ng mostly associated with particles >1.5 mm (Tables 3 – 5). Higher mÀ3; Table 5) aerosol. High concentrations of n-alkenoic acids concentrations of n-alkanoic acids, associated with particles >1.5 were associated with large particles in forest samples (e.g., >3.0 mm, were measured in the forest atmosphere (from 9.9 to 44.0 ng mm, 3.5 – 6.4 ng mÀ3; Table 5). The particle size distribution pattern mÀ3; Table 5) than the other sites (urban, 0.2 – 0.9 ng mÀ3 (Table 3); of n-alkenoic acids in the urban area suggests their association with background marine, 1.7 – 4.3 ng mÀ3 (Table 4)). The n-alkanoic acid particles <1.5 mm (Figure 4b). Conversely, the corresponding concentration levels in fine and ultrafine fractions (<1.5 mm) were Lundgren diagram for marine and forest aerosol demonstrated a considerably higher for urban (Table 3) and forest (Table 5) bimodal profile (Figure 4b). These differences suggest different aerosols. The enrichment of n-alkanoic acids at particles <1.5 mm sources for n-alkenoic acids among the sampling sites. Emissions was observed in the urban environment (Figure 4a and Table 3).
of n-alkenoic acids from higher terrestrial plants in the forest were This was also true for a large fraction of n-alcanoic acids in marine associated with large particles. Coagulation and agglomeration and forest aerosol (Figure 4a and Tables 4 and 5); however, a second between particles along with local emissions of large particles maximum was also observed for larger particles (>7.2 mm).
may be the major source of n-alkenoic acids in the background [19] The strong even-to-odd predominance expressed by the high CPI values suggests that direct emission from epicuticular [21] The photooxidation products of unsaturated acids, a, waxes of higher terrestrial plants was the predominant source in w-dicarboxylic acids, were detected in ambient sample in all sites urban (CPI 10.6 – 19.1; Table 3), marine (CPI 11.1 – 24.5; Table 4), (Tables 1 and 3 – 5). In all samples, a, w-dicarboxylic acids ranged and forest (5.0 – 8.8; Table 5) aerosol. These results imply that from C6 to C15 (Tables 3 – 5). The homologue distribution maxi- biogenic organic acids may be associated with both fine and coarse mized at azelaic (C9) acid. The C5 and C6 homologues are formed particles. In general, n-alkanoic acids originate from a variety of by the oxidation of cyclic olefins [Stephanou and Stratigakis, sources, and therefore their specific origin reconciliation should be 1993] and other emission sources [Rogge et al., 1993a, 1993b, cautious [Simoneit, 1999; Rogge et al., 1993a, 1993b, 1993c, 1993c, 1994, 1998], while the C8 and C9 ones (the most abundant; see Tables 3 – 5) are formed by photooxidation of unsaturated [20] A series of monounsaturated (n-alkenoic) acids were carboxylic acids such as oleic (C18:1) and linoleic (C18:2) acids detected in ambient samples. Palmitoleic (C16:1) and oleic (C18:1) [Stephanou, 1992; Stephanou and Stratigakis, 1993]. Since oleic acids were detected in all samples (Tables 3 – 5). Higher molecular (C18:1) and linoleic (C18:2) acids are products of atmospheric weight unsaturated acids (up to C20:1; Table 4) were detected in reactions in both gas and particle phase [Stephanou, 1992; background marine aerosol. In addition, pentadecenoic acid (C15:1) Stephanou and Stratigakis, 1993], they were mainly detected in was identified in forest aerosol (Table 5). This variability can be the fine and ultrafine particles (<1.5 mm) in all areas (Figure 4c explained by the contribution of other sources to marine aerosol such as phytoplankton [Gogou et al., 1996]. The concentrations of [22] Finally, series of carbonyl and carboxyl compounds were n-alkenoic acids measured in all impactor stages in the marine area detected in forest samples. The chemical structure of these com- KAVOURAS AND STEPHANOU: SIZE DISTRIBUTION OF ORGANIC AEROSOL pounds was identified by their mass spectra [Kavouras et al., We thank N. Stratigakis and N. Mihalopoulos for their assistance during 1999b]. These compounds were pinonic acid, pinic acid, pinonal- dehyde and nopinone, and are considered as photooxidationproducts of a- and b-pinene [Kavouras et al., 1999a, 1999b]. It was recently proven that these compounds were responsible for the Aboulkassim, T. A. T., and B. R. T. Simoneit, Lipid geochemistry of formation of secondary organic aerosol over forests [Kavouras et surficial sediments from the coastal environment of Egypt, 1, Aliphatic al., 1998b, 1999a, 1999b]. As it is shown in Lundgren diagrams hydrocarbons-characterization and source, Mar. Chem., 54, 135 – 158,1996.
(Figure 5) both carbonyl and carboxyl compounds were primarily Aceves, M., and J. O. Grimalt, Seasonally dependent size distributions of associated with particles <0.96 mm. More than 90% of pinonic and aliphatic and polynuclear hydrocarbons in urban aerosols from densely pinic acid and nopinone was associated with particles <0.96 mm populated areas, Environ. Sci. Technol., 27, 2896 – 2908, 1993.
(Table 5). A high amount of pinonaldehyde was associated with Allen, J. O., N. M. Dookeran, K. Taghizadeh, A. L. Lafleur, K. A. Smith, particles with a diameter from 0.5 to 0.96 mm. These results and A. F. Sarofim, Measurement of oxygenated polycyclic aromatic hy- suggest these compounds may be responsible for the higher aerosol drocarbons associated with a size-segregated urban aerosol, Environ. Sci.
Technol., 31, 2064 – 2070, 1997.
and organic extract mass in smaller particle sizes (Table 2).
Andreae, M. O., R. W. Talbot, T. W. Andreae, and R. C. Harris, Formic [23] The comparison of MMAD, calculated for the total particle and aceic acid over the central Amazon region, Brazil, 1, Dry season, size range and for fine particles, between n-alkanoic (0.71, 0.62, J. Geophys. Res., 92, 6635 – 6641, 1988.
and 0.91 mm; Table 6), n-alkenoic (0.86, 1.65, and 1.22 mm; Table Ansari, A. S., and S. N. Pandis, Water absorption by secondary organic 6), dicarboxylic acids (0.34, 0.77, 0.34 mm; Table 6), and mono- aerosol and its effect on inorganic aerosol behavior, Environ. Sci. Technol.,34, 71 – 77, 2000.
terpene photooxidation products (0.28 – 0.66 mm; Table 6) reflects Atkinson, R., Gas-phase troposperic chemistry of organic compounds: the processes of aerosol formation for these compound classes. A A review, Atmos. Environ., 24A, 1 – 41, 1990.
secondary process is responsible for the occurrence of dicarboxylic Atkinson, R., and J. Arey, Atmospheric chemistry of biogenic organic acids and monoterpene photooxidation products, while direct compounds, Account. Chem. Res., 31, 574 – 583, 1998.
emission is responsible for the presence of n-alkanoic and n- Chow, J. C., J. G. Watson, D. H. Lowenthal, P. A. Solomon, K. L.
alkenoic acids. However, our results indicate that the role of higher Magliano, S. D. Ziman, and L. W. Richards, PM10 and PM2.5 composi-tions in California’s San Joaquin Valley, Aerosol Sci. Technol., 18, molecular weight carbonyl and carboxyl compounds might also be important for the formation of ultrafine particles in the atmosphere.
Chow, J. C., J. G. Watson, E. M. Fujita, Z. Q. Lu, D. R. Lawson, and L. L.
Ashbaugh, Temporal and spatial variations of PM2.5 and PM10 aerosol inthe Southern California Air Quality Study, Atmos. Environ., 28, 2061 – Cruz, C. N., and S. N. Pandis, The effect of organic coatings on the [24] The size distribution of organic aerosol in urban, marine cloud condensation nuclei activation of inorganic atmospheric aerosol, background, and forest areas was characterized on the basis of J. Geophys. Res., 103, 13,111 – 13,123, 1998.
MMAD calculation and Lundgren diagram construction. Most of Fernandes, M. B., M. A. Sicre, I. Broyelle, A. Lorre, and D. Pont, Con- the extractable organic (up to 85%) matter was associated in all tamination by polycyclic aromatic hydrocarbons (PAHs) in French and cases with fine particles (<1.5 mm). Urban aerosol extractable European rivers, Hydrobiologia, 410, 343 – 348, 1999.
Gogou, A., N. Stratigakis, M. Kanakidou, and E. G. Stephanou, Organic organic mass was distributed between fine particles with a MMAD aerosols in Eastern Mediterranean: Components source reconciliation by of 0.60 mm and coarse particles with a MMAD of 5.59 mm. For the using molecular markers and atmospheric back trajectories, Org. Geo- forest aerosol, extractable organic mass was distributed between fine particles with MMAD of 0.36 mm and coarse particles with a Gogou, A., M. Apostolaki, and E. G. Stephanou, Determination of organic MMAD of 5.13 mm. Organic molecular markers characterized in molecular markers in marine aerosols and sediments: One-step flash the three different aerosol types have shown that size distribution chromatography compounds class fractionation and capillary gaschromatographic analysis, J. Chromatogr. A, 799, 215 – 231, 1998.
was primarily related to the emission source of these compounds.
Gogou, A., I. Bouloubassi, and E. G. Stephanou, Marine organic geochem- In addition, the effect of coagulation and nucleation might be istry of the Eastern Mediterranean, 1, Aliphatic and polyaromatic hydro- significant in specific areas (e.g., urban environment). In particular, carbons in Cretan Sea surficial sediments, Mar. Chem., 68, 265 – 282, n-alkanes were mostly associated with fine particles in the urban and forest aerosols, and their total MMAD was 0.45 mm and 0.63 Hahn, J., Organic constituents of natural aerosol, Ann. N. Y. Acad. Sci., 338, mm, respectively. In the background marine aerosol, n-alkanes were Jaenicke, R., The role of organic material in atmospheric aerosols, Pure more evenly distributed, and their MMAD was 2.00 mm, showing Appl. Geophys., 116, 283 – 292, 1978.
physical changes due to long-range transport. Similar observations Kavouras, I. G., N. Stratigakis, and E. G. Stephanou, Iso- and anteiso- have been done for PAHs and n-alkanals. Conversely, the most alkanes: Specific tracers of environmental tobacco smoke in indoor and biogenic compound class, namely n-alkanols, were evenly asso- outdoor particle-size distributed urban aerosols, Environ. Sci. Technol., ciated in the urban, background marine, and forest aerosols, Kavouras, I. G., N. Mihalopoulos, and E. G. Stephanou, Formation of between fine and coarse particles, and their corresponding total atmospheric particles from organic acids produced by forests, Nature, MMADs were 2.45, 2.69, and 1.67 mm, respectively. Saturated and unsaturated acids were mainly concentrated in the fine fraction; Kavouras, I. G., N. Mihalopoulos, and E. G. Stephanou, Secondary organic however, high concentrations of acids were detected in the coarse aerosol formation vs. primary organic aerosol emission: In situ evidence fraction of forests. The total MMADs of n-alkanoic acids were for the chemical coupling between monoterpene acidic photooxidation 0.71, 0.62, and 0.91 mm in the urban, background marine, and products and new particle formation over forests, Environ. Sci. Technol.,33, 1028 – 1037, 1999a.
forest aerosols, respectively. Several low molecular weight organic Kavouras, I. G., N. Mihalopoulos, and E. G. Stephanou, Formation and gas/ compounds associated with photochemical reactions in the atmos- particle partitioning of monoterpenes photo-oxidation products over for- phere were detected in urban marine and forests aerosols. Specif- ests, Geophys. Res. Lett., 26, 55 – 58, 1999b.
ically, nonanal (urban), azelaic acid (urban and marine), pinonic Lowenthal, D. H., B. Zielinska, J. C. Chow, J. G. Watson, M. Gautam, acid, pinonaldehyde, pinic acid, and nopinone (forest) were found D. H. Ferguson, G. R. Neuroth, and K. D. Stevens, Characterization ofheavy duty diesel vehicle emissions, Atmos. Environ., 28, 731 – 743, in the fine and ultrafine fraction showing the low total MMAD (0.28 – 0.77 mm) in the aerosol types studied.
Mihalopoulos, N., E. G. Stephanou, M. Kanakidou, S. Pilitsidis, and P. Bousquet, Tropospheric aerosol composition in the Eastern Med-iterranean region, Tellus, Ser. B, 49, 314 – 326, 1997.
[25] Acknowledgments. European Commission (DGXII, Environ- Pagano, P., T. Zalamaco, E. Scarcella, S. Bruni, and M. Calamosca, Muta- ment and Climate Program, ENV4-CT95-0049) and the Special Research genic activity of total particle-sized fractions of urban particulate matter, Account of the University of Crete are acknowledged for financial support.
Environ. Sci. Technol., 30, 3512 – 3516, 1996.
KAVOURAS AND STEPHANOU: SIZE DISTRIBUTION OF ORGANIC AEROSOL Pandis, S. N., R. A. Harley, G. R. Cass, and J. H. Seinfeld, Secondary hydrocarbons over the Mediterranean Sea, Geophys. Res. Lett., 17, organic aerosol formation and transport, Atmos. Environ., 26, 2269 – Sicre, M. A., J. C. Marty, and A. Saliot, n-Alkanes, fatty acid esters and Pankow, J. F., Partitioning of semi-volatile organic compounds to the air/ fatty acid salts in size fractionated aerosol collected over the Mediterra- water interface, Atmos. Environ., 31, 927 – 929, 1997.
nean Sea, J. Geophys. Res., 95, 3649 – 3657, 1990b.
Rogge, W. F., L. Hildemann, M. A. Mazurek, G. R. Cass, and B. R. T.
Simoneit, B. R. T., A review of biomarker compounds as source indicators Simoneit, Sources of fine organic aerosol, 2, Noncatalyst and catalyst- and tracers of air pollution, Environ. Sci. Pollut., 6, 159 – 169, 1999.
equipped automobiles and heavy duty diesel trucks, Environ. Sci. and Stephanou, E. G., Long chain n-aldehydes: An ubiquitous compound class of environmental and geochemical significance, Naturwissenschaften, 76, Rogge, W. F., L. Hildemann, M. A. Mazurek, G. R. Cass, and B. R. T.
Simoneit, Sources of fine organic aerosol, 3, Road dust, tire debris and Stephanou, E. G., a, w-Dicarboxylic acid salts and a, w-dicarboxylic acids: organometallic brake lining dust: Roads as sources and sinks, Environ.
Photo-oxidation products of unsaturated fatty acids, present in marine Sci. Technol., 27, 1892 – 1904, 1993b.
aerosols and marine sediments, Naturwissenschaften, 79, 128 – 131, Rogge, W. F., L. Hildemann, M. A. Mazurek, G. R. Cass, and B. R. T.
Simoneit, Sources of fine organic aerosol, 4, Particulate abrasion products Stephanou, E. G., and N. Stratigakis, Oxocarboxylic and a, w-dicarboxylic from leaf surfaces of urban plants, Environ. Sci. Technol., 27, 2700 – acids: Photooxidation products of biogenic unsaturated fatty acids present in urban aerosols, Environ. Sci. Technol., 27, 1403 – 1407, 1993.
Rogge, W. F., L. Hildemann, M. A. Mazurek, G. R. Cass, and B. R. T.
Talbot, R. W., A. S. Vijgen, and R. C. Harriss, Soluble species in the Arctic Simoneit, Sources of fine organic aerosol, 5, Natural gas home appli- summer troposphere: Acidic gases, aerosol and precipitation, J. Geophys.
ances, Environ. Sci. Technol., 27, 2736 – 2744, 1993d.
Rogge, W. F., L. Hildemann, M. A. Mazurek, G. R. Cass, and B. R. T.
Tao, Y., and P. H. McMurry, Vapor pressure and surface free energies of Simoneit, Sources of fine organic aerosol, 6, Cigarette smoke in the urban C14 – C18 monocarboxylic acids and C5 and C6 dicarboxylic acids, atmosphere, Environ. Sci. Technol., 28, 1375 – 1388, 1994.
Environ. Sci. Technol., 23, 1519 – 1523, 1989.
Rogge, W. F., L. Hildemann, M. A. Mazurek, G. R. Cass, and B. R. T.
Turpin, B. J., P. Saxena, and E. Andrews, Measuring and simulating Simoneit, Sources of fine organic aerosol, 7, Hot asphalt roofing tar pot particulate organics in the atmosphere: Problems and prospects, Atmos.
fumes, Environ. Sci. Technol., 31, 2726 – 2730, 1997a.
Rogge, W. F., L. Hildemann, M. A. Mazurek, G. R. Cass, and B. R. T.
Van Vaeck, L., and K. Van Cauwenberghe, Characteristic parameters of Simoneit, Sources of fine organic aerosol, 8, Boilers burning distillate particle size distributions of primary organic constituents of ambient fuel oil, Environ. Sci. Technol., 31, 2731 – 2737, 1997b.
aerosols, Environ. Sci. Technol., 19, 707 – 716, 1985.
Rogge, W. F., L. Hildemann, M. A. Mazurek, G. R. Cass, and B. R. T.
Venkataraman, C., and S. K. Friedlander, Size-distribution of polycyclic Simoneit, Sources of fine organic aerosol, 9, Pine, oak and synthetic log aromatic hydrocarbons and elemental carbon, 2, Ambient measurements combustion in residential fireplaces, Environ. Sci. Technol., 32, 13 – 22, and effects of atmospheric processes, Environ. Sci. Technol., 28, 563 – Russell, L. M., S. N. Pandis, and J. H. Seinfeld, Aerosol production and growth in the marine boundary layer, J. Geophys. Res., 99, 20,989 –21,003, 1994.
Sicre, M. A., J. C. Marty, A. Saliot, X. Aparicio, J. Grimalt, and J. Albaiges, Aliphatic and aromatic hydrocarbons in different sized aerosols over the Mediterranean Sea: Occurrence and origin, Atmos. Environ., 21, 2247 – I. G. Kavouras and E. G. Stephanou, Environmental Chemical Processes Laboratory, Department of Chemistry, University of Crete, GR-71409 Sicre, M. A., J. C. Marty, A. Lorre, and A. Saliot, Airborne and vapor-phase Heraklion, Hellas, Greece. ([email protected])

Source: http://finokalia.chemistry.uoc.gr/data/free/publications/Kavouras%20I.G.,%20%20and%20E.G.%20Stephanou,%202002.pdf