Field performance of dichotomous sequential PM air samplers

For over one year, the Environmental Protection Commission of Hillsborough County (EPCHC) in Tampa, Florida, operated two dichotomous sequential particulate matter air samplers collocated with a manual Federal Reference Method (FRM) air sampler at a waterfront site on Tampa Bay. The FRM was alternately configured as a PM_2.5, then as a PM₁₀ sampler. For the dichotomous sampler measurements, daily 24-h integrated PM_2.5 and PM_10–2.5 ambient air samples were collected at a total flow rate of 16.7 l min⁻¹. A virtual impactor split the air into flow rates of 1.67 and 15.0 l min⁻¹ onto PM_10–2.5 and PM_2.5 47-mm diameter PTFE^® filters, respectively. Between the two dichotomous air samplers, the average concentration, relative bias and relative precision were 13.3 μg m⁻³, 0.02% and 5.2% for PM_2.5 concentrations (n=282), and 12.3 μg m⁻³, 3.9% and 7.7% for PM_10–2.5 concentrations (n=282). FRM measurements were alternate day 24-h integrated PM_2.5 or PM₁₀ ambient air samples collected onto 47-mm diameter PTFE^® filters at a flow rate of 16.7 l min⁻¹. Between a dichotomous and a PM_2.5 FRM air sampler, the average concentration, relative bias and relative precision were 12.4 μg m⁻³, −5.6% and 8.2% (n=43); and between a dichotomous and a PM₁₀ FRM air sampler, the average concentration, relative bias and relative precision were 25.7 μg m⁻³, −4.0% and 5.8% (n=102). The PM_2.5 concentration measurement standard errors were 0.95, 0.79 and 1.02 μg m⁻³; for PM₁₀ the standard errors were 1.06, 1.59, and 1.70 μg m⁻³ for two dichotomous and one FRM samplers, respectively, which indicate the dichotomous samplers have superior technical merit. These results reveal the potential for the dichotomous sequential air sampler to replace the combination of the PM_2.5 and PM₁₀ FRM air samplers, offering the capability of making simultaneous, self-consistent determinations of these particulate matter fractions in a routine ambient monitoring mode.     

Iron,manganese and copper emitted by cargo and passenger trains in Zürich (Switzerland): Size-segregated mass concentrations in ambient air

Particle emissions caused by railway traffic have hardly been investigated in the past, due to their obviously minor influence on air quality compared to automotive traffic. In this study, emissions related to particle abrasion from wheels and tracks were investigated next to a busy railway line in Zürich (Switzerland), where trains run nearly exclusively with electrical locomotives. Hourly size-segregated aerosol samples (0.1–1, 1–2.5 and 2.5–10 μm) were collected with a rotating drum impactor (RDI) and subsequently analyzed by synchrotron radiation X-ray fluorescence spectrometry (SR-XRF). In this way, hourly elemental mass concentrations were obtained for chromium, manganese, iron and copper, which are the elements most relevant for railway abrasion. Additionally, daily aerosol filters were collected at the same site as well as at a background site for subsequent analysis by gravimetry and wavelength dispersive XRF (WD-XRF). Railway related ambient air concentrations of iron and manganese were calculated for the coarse (2.5–10 μm) and fine (<2.5 μm) particle fraction by means of a Mn/Fe ratio investigation. The comparison to train type and frequency data showed that 75% and 60% of the iron and manganese mass concentrations related to cargo and passenger trains, respectively, were found in the coarse mode. The railway related iron mass concentration normalized by the train frequency ranges between 10 and 100 ng m⁻³ h iron in 10 m distance to the tracks, depending on train type. It is estimated that the personal exposure next to a busy railway line above ground is more than a magnitude lower than inside a subway station.     

Seasonal characteristics of water-soluble organic carbon in atmospheric particles in the inland Kanto plain,Japan

An investigation of water-soluble organic carbon (WSOC) in atmospheric particles was conducted as an index of the formation of secondary organic aerosol (SOA) from April 2005 to March 2006 at Maebashi and Akagi located in the inland Kanto plain in Japan. Fine (<2.1 μm) and coarse (2.1–11 μm) particles were collected by using an Andersen low-volume air sampler, and WSOC, organic carbon (OC), elemental carbon (EC), and ionic components were measured. The mean mass concentrations of the fine particles were 22.2 and 10.5 μg m^?3 at Maebashi and Akagi, respectively. The WSOC in fine particles accounted for a large proportion (83%) of total WSOC. The concentration of fine WSOC ranged from 1.2 to 3.5 μg-C m^?3 at Maebashi, rising from summer to fall. At Akagi, it rose from spring to summer, associated with the southerly wind from urban areas. The WSOC/OC ratio increased in summer at both sites, but the ratio at Akagi was higher, which we attributed to differences in primary emissions and secondary formation between the sites. The fine WSOC concentration was significantly positively correlated with concentrations of SO₄^2?, EC, and K⁺, and we inferred that WSOC was produced by photochemical reaction and caused by the combustion of both fuel and biomass. We estimated that SOA accounted for 11–30% of the fine particle mass concentration in this study, suggesting that SOA is a significant year-round component in fine particles.     

Verification of traffic emitted aerosol components in the ambient air of Cologne (Germany)

Emission of heavy metals, besides ozone and diesel-soot, is one of the most significant environmental problems caused by the existing transport systems. Emission arises from different parts of vehicles (tyres, brakes, exhausts) or running trains (rails, wheels, overhead cables, etc.). Various types of emission with their insufficiently known ways of spreading makes it difficult to estimate the risk and to realise effective counter-measures.Size-dependent sampling of aerosols with the virtual impactor technique of a dichotomous sampler (<2.5 μm, 2.5–10 μm) allows one to identify several aerosol components at the sampling locations and to reliably determine their relative proportions. Analyses of 30–40 elements from immission samplings at 12 places in and around Cologne are compared with specific patterns of elements from relevant emission sources. Moreover, taking into account the particle size, relative abundances of Zn, Mo, Cu and Sb can be estimated by assuming an average of around 30% of diesel-soot in the fine fraction, and ∼3% from an abraded tyre material together with ∼0.3% from rubbed off brake-linings in the coarse fraction. The assumption of 0.3% for brake-linings is based on the relative abundances of Cu and Sb which besides Cd are the most highly enriched. Since most brake-linings used in cars contain these elements in an unusually high quantity (5–20% Cu and 1–5% Sb) and are in very similar ratios as measured in the coarse fraction, Cu and Sb may be taken as quantitative tracers for the brake-lining component in the immission. The environmental interest in Sb arises mainly from the toxicological potential of the compounds Sb₂S₃ and Sb₂O₃.Other traffic related components could only be identified very close to the sources of emissions. Besides Pt from cars with catalysators and Cu emitted from overhead cables of trams, an As-enrichment from rusting rails, which segregates into fine particles because of shaking due to passing trains, was discovered.     

Inorganic element concentrations in near surface aerosols sampled on the northwest slopes of Mount Kenya

Eight trace elements, Si, Cl, K, Ca, Ti, Mn, Fe and Zn in the near-ground atmospheric aerosols were evaluated in the northwestern part of Mount Kenya using a dichotomous sampler and an EDXRF spectrometer. The samples were taken at 2 sites situated in Nanyuki area, which is roughly on the Equator. The sampler segregated the aerosol into two aerodynamic diameter (ad) size fractions, fine (<3.5 μm ad) and coarse (>3.5 and <18 μm ad). The elemental concentrations in the two size fractions were quantified and the elements assigned to known sources. Local wind blown dust related to agricultural activities and fire burning was found to dominate the lower tropospheric aerosols. There was inconclusive evidence of long range-transported aerosols being moved by night transport from the middle to the lower parts of the troposphere. Influence of the Indian Ocean marine aerosol was suggested but conclusive evidence was lacking.     

Size distribution of metals in urban aerosols in Seville (Spain)

The size distribution of metals in aerosols has been studied in 12 areas of the city of Seville. Urban particles were collected with a high-volume sampling system equipped with a cascade impactor, which effectively separates the particulate matter into six-size ranges. Forty-one samples were collected in spring 1996. Each filter was extracted with a mixture of nitric and percloric acids. The acid solutions of the samples were analysed in six-particle fractions by inductively coupled plasma atomic emission spectrometry (ICP-AES). The impactor stage fractionation of particles shows a typical bimodal distributions, one corresponding to the fine mode below 1 μm (55%), and the other to the coarse mode around 10 μm (32%). With regard to the size distribution of metals, we concluded that potentially toxic metals, such as nickel, lead and cadmium are mainly accumulated in the smaller particles, with percentages of 72.6, 69.4 and 63.8%, respectively. Lead have a concentration of 63.7 ng m⁻³, more than copper and manganese (26.7 and 16.5 ng m⁻³) and above all more than nickel, cobalt and cadmium (1.97, 0.54 and 0.32 ng m⁻³).     

Coarse atmospheric aerosol: size distributions of trace elements

A sampler, employing nine single stage impactors placed in parallel within a portable wind tunnel, has been used to determine the metal content of coarse atmospheric aerosol. The wind tunnel maintains a constant flow environment for the collectors housed inside it, so that representative sampling conditions are achieved compared to the varied ambient wind conditions. At a flow rate of 8 m s⁻¹ the 50% cut-off diameters of the impactors ranged from 7.8 to 38.8 μm. Measurements were conducted at a rural and urban site near Colchester in south east England. The samplers were analysed by PIXE for P, K, Ca, Fe, Ti, Mn, Cu, V, Co, Cr, Br, Zn, Ni, Sc and Pb. It is found that the sampler can be employed to quantitatively characterise the elemental mass size distribution for aerosol larger than 10 μm. The results indicate that a small fraction of the above earth and trace elements’ metal mass is present in particles greater than 10 μm. This fraction for earth metals (Ca, K, Ti) is comparatively greater in the rural site than the urban site, while for trace metals (Mn, V, Cu, Cr) this fraction constitutes a more significant part of the coarse mass at the urban site. Trace element concentrations were of a similar order of magnitude to earlier literature reports. Although the number of measurements was limited it can be concluded that the size distributions obtained were characteristic of an unpolluted area.     

Characterisation of PM10 atmospheric aerosols for the wet season 2005 at two sites in East Africa

Ambient daily PM₁₀ aerosol samples were collected at two sites in Tanzania in May and June 2005 (during the wet season), and their chemical characteristics were studied. The sites were a rural site in Morogoro and an urban kerbside site in Dar es Salaam. A Gent PM₁₀ stacked filter unit sampler with sequential Nuclepore polycarbonate filters, providing fine and coarse size fractions, and a PM₁₀ sampler with quartz fibre filters were deployed. Parallel collections of 24 h were made with the two samplers and the number of these collections was 13 in Morogoro and 16 in Dar es Salaam. The average mass concentration of PM₁₀ was 27 ± 11 μg/m³ in Morogoro and 51 ± 21 μg/m³ in Dar es Salaam. In Morogoro, the mean concentrations of organic carbon (OC), elemental carbon (EC), and water-soluble organic carbon (WSOC) were 6.8, 0.51, and 2.8 μg/m³, respectively. In contrast, higher mean concentrations (11.9, 4.6, and 3.3 μg/m³, respectively) were obtained for Dar es Salaam. At both sites, species and elements, such as black carbon, NH₄⁺, non-sea-salt SO₄^2?, K, and Ni (and at Dar es Salaam also V, As, Br, and Pb) were mainly present in the fine size fraction. The common crustal and sea-salt elements, including Na, Mg, Al, Si, Cl, Ca, Ti, Mn, Fe, and Sr, and also NO₃^? and P (and to a lesser extent Cu and Zn) were concentrated in the coarse particles. Aerosol chemical mass closure indicated that the PM₁₀ mass in Morogoro consisted, on average, of 48% organic matter (OM), 44% crustal matter, 4% sea salt, and 2% EC, while in Dar es Salaam OM, crustal matter, sea salt, and EC represented 37%, 32%, 9%, and 9% of the PM₁₀ mass. The contributions of the secondary inorganic aerosol (non-sea-salt sulphate, nitrate, and ammonium) were small, i.e., only 5% in total at each site. Carbonaceous materials and crustal matter were thus the most important components of the PM₁₀ mass. It is suggested that biomass burning is a major contributor to the OM; at Dar es Salaam there is also a very substantial contribution from traffic. A source apportionment calculation indicated that 68% of the OC at this site originated from traffic exhaust versus 32% from charcoal burning. The crustal matter at Morogoro is likely mainly attributable to soil dust resuspension, whereas in Dar es Salaam it is likely mostly resuspended road dust.     

Study of size distribution of atmospheric aerosol at Agra

Measurements on size distribution of atmospheric aerosol were made at Dayalbagh, Agra during July to September 1998. A 4-stage cascade particle sampler (CPS - 105) which fractionates particles in sizes ranging between 0.7 and >10.9 μm, was used. Samples were collected on Whatman 41 filters. The filters were analyzed for the major water-soluble ions. The anions (F, Cl, NO₃ and SO₄) were analyzed by Dionex DX-500 ion chromatograph while atomic absorption and colorimetric techniques were used for the analysis of cations (Na, K, Ca and Mg) and NH₄, respectively. The average mass of aerosol was found to be 131.6 μg m⁻³ and aerosol composition was found to be influenced by terrigeneous sources. The mass size distribution of total aerosol and the ions NH₄, Cl, NO₃, K, Ca, Mg, SO₄ and Na was bimodal while that of F was unimodal. SO₄, F, K and NH₄ dominated in the fine mode while Ca, Mg, Cl and NO₃ were in abundance in coarse fraction. Na was found in both coarse as well as fine mode. Coarse mode SO₄ and NO₃ have been ascribed to contribution from re-suspension of soil and formation by heterogeneous oxidation on soil derived particles. Preponderance of K in fine mode is attributed to emissions from vegetation and from burning of plant materials. Ca, Mg, Cl and NO₃ are largely soil derived and hence dominate in coarse fraction. Equivalent ratios of NH₄/(SO₄+NO₃) were calculated for both fine and coarse aerosols. The coarse mode ratio varied between 0.7 and 1.4 while in fine mode it ranged between 1.4 and 1.9. It shows that aerosol is basic, the basicity of coarse mode is due to higher concentration of soil-derived alkaline components while the basicity in fine mode is due to neutralization of acidity by NH₃.     

High volume electrostatic field-sampler for collection of fine particle bulk samples

A high volume electrostatic field-sampler was developed for collection of fine particles, which easily can be recovered for subsequent sample characterisation and bioassays. The sampler was based on a commercial office air cleaner and consisted of a prefilter followed by electrostatic collection plates operating at 2.7 kV. The sampler performance was characterised for 26 nm to 5.4 μm-size particles in urban street air. The collection efficiency reached a maximum (60–70%) between 0.2 and 0.8 μm and dropped to ∼25% at 30 nm and 2.5 μm, respectively. After extraction in water, the particle loss was<2%. The extraction efficiency for dry lyophilised particulate matter was above 80%, allowing retrievement of ∼12 mg day⁻¹ in urban street air at PM₁₀ levels of ∼24 μg m⁻³. The ozone generating capacity of the corona discharge during operation was on the order of 10 ppb. A polycyclic aromatic hydrocarbons (PAH) degradation test using benzo[a]pyrene as a model showed that ∼85% was degraded after 24 h. However, similar results were observed when the corona discharge was switched off. Hence, the ozone and other corona discharge reactants do not appear to contribute considerably to PAH-degradation. The overall results show that the sampler type is a promising alternative to traditional sampling of fine particles for bulk analysis and bioassays. The main advantages are simple operation, high stability, high quantifiable particle recovery rates and low cost.     

Distribution characteristics of polycyclic aromatic hydrocarbons with particle size in urban aerosols at the roadside in Ho Chi Minh City,Vietnam

The purpose of this study was to characterize size distributions of atmospheric polycyclic aromatic hydrocarbons (PAHs) with 4–6 rings at the roadside in Ho Chi Minh City, Vietnam. Ten PAHs (fluoranthene, pyrene, triphenylene, benzo[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[a]pyrene, benzo[ghi]perylene and indeno[1,2,3-cd]pyrene) in atmospheric particulate matters (PM) at the roadside were measured in the dry and rainy seasons in 2005 at Ho Chi Minh City, using a low-pressure cascade impactor. The PM were separated into nine fractions by their aerodynamic diameter, i.e. >9.0, 9.0–5.8, 5.8–4.7, 4.7–3.3, 3.3–2.1, 2.1–1.1, 1.1–0.7, 0.7–0.4 and <0.4 μm (a final filter). PAHs were analyzed by high-performance liquid chromatography with fluorescence detection. Total PAHs measured were higher in the rainy season than in the dry season. The mass of coarse particles occupied a higher fraction than that of fine particles in both seasons. Total PAHs were mainly concentrated in particles with aerodynamic diameter smaller than 0.4 μm. The particle size distributions of PAHs investigated were bi-modal with a peak in fine particle mode (<2.1 μm) and another peak in coarse particle mode (>2.1 μm). Generally, 5,6-ring PAHs associated mainly with fine particles and 4-ring PAHs spread out in both fine and coarse particles.     

Studies of the coarse particle (2.5–10 μm) component in UK urban atmospheres

An analysis is presented of continuous simultaneous measurement data for PM₁₀ and PM_2.5 using TEOM instruments from five sites in the United Kingdom. The results are analysed specifically in relation to the sources and processes influencing the coarse particle fraction (2.5–10 μm). The data show a generally strong correlation between fine and coarse particle concentrations at all sites, with a generally higher proportion of coarse particles in the dryer months of the year. The one rural site shows a notably lower proportion of coarse particles than the urban and suburban sites. Whilst it is possible to disaggregate the coarse particle concentrations into a component which is diluted by increasing windspeed and a component which increases with windspeed and is hence possibly attributable to wind-induced resuspension processes, the latter is only a minor proportion of the total coarse particle concentration. There are appreciable weekday-to-weekend and day-to-night differences between coarse particle concentrations which are most marked at the urban sites indicative of anthropogenic activities being a source of coarse particles. The clearest indication of the likely predominant source of coarse particles arises from an analysis of a data set derived from an urban street canyon site after subtraction of measurements from a nearby urban background location. The data indicate strong relationships of both fine and coarse incremental particle concentrations in the street canyon with incremental NO_x. If incremental fine particles and coarse particles are attributed to exhaust emissions and vehicle-induced resuspension, respectively, then it may be concluded that vehicle-induced resuspension provides a source strength approximately equal to that of exhaust emissions. An analysis of the coarse particle concentration data suggest that episodes of elevated coarse particle concentrations alone very rarely lead to exceedence of the UK air quality standard for PM₁₀ of 50 μg m⁻³ measured as a 24-h running mean.     

Experimental studies on particle impaction and bounce: effects of substrate design and material

This paper presents an experimental investigation of the effects of impaction substrate designs and material in reducing particle bounce and reentrainment. Particle collection without coating by using combinations of different impaction substrate designs and surface materials was conducted using a personal particle sampler (PPS) developed by the University of Southern California. The PPS operates at flow rate of 4 l min^-1 with a 50% cutpoint of approximately 0.9 μm in aerodynamic diameter. The laboratory results showed that the PPS collection efficiency for particles larger than 50% cutpoint is strikingly low (e.g., less than 50%) when an uncoated open cavity made of aluminum was used as an impaction substrate. The collection efficiency gradually increased when Teflon tape, Nuclepore, and glass fiber filters were used as impaction surfaces, respectively. Conical or partially enclosed cavity substrate designs increased collection efficiency of particles of 9 μm up to 80–90%. A conical cavity with glass fiber filter used as impaction surface was identified as the optimum configuration, resulting in a collection efficiency of 92% at Stokes numbers as high as 15.4 (corresponding to 9 μm in aerodynamic diameter). Particle losses were low (less than 10%) and relatively independent of particle size in any design with glass fiber filter. Losses seemed to increase slightly with particle size in all other configurations. Finally, outdoor PM₁ concentrations obtained with the PPS (in its optimum configuration) and a modified micro-orifice uniform deposit impactor (MOUDI) with coated impaction stages were in excellent agreement. The mean ratio of the PPS-to-MOUDI concentration was 1.13(±0.17) with a correlation coefficient R²=0.95. Results from this investigation can be readily applied to design particle bounce-free impaction substrates without the use of coating. This is a very important feature of impactors, especially when chemical analysis of the collected particulate matter is desirable.     

Quantitative analysis of volatile organic compounds (VOCs) in atmospheric particles

There are a number of difficulties associated with the quantitative analysis of volatile organic compounds (VOCs) in atmospheric particles. Therefore, majority of the previous studies on VOCs associated with particles have been qualitative. Air samples were collected in Izmir, Turkey to determine ambient particle and gas phase concentrations of several aromatic, oxygenated and halogenated VOCs. Samples were quantitatively analyzed using thermal desorption–gas chromatography/mass spectrometry. Gas-phase concentrations ranged between 0.02 (bromoform) and 4.65 μg m⁻³ (toluene) and were similar to those previously measured at the same site. Particle-phase concentrations ranged from 1 (1,3-dichlorobenzene) to 933 pg m⁻³ (butanol). VOCs were mostly found in gas-phase (99.9±0.25%). However, the particulate VOCs had comparable concentrations to those reported previously for semivolatile organic compounds. The distribution of particle-phase VOCs between fine (d_p<2.5 μm) and coarse (2.5 μm<d_p<10 μm) fractions was also investigated. It was found that VOCs were mostly associated with fine particles.     

Physicochemical characteristics of atmospheric aerosol during winter in the São Paulo Metropolitan area in Brazil

In order to characterize atmospheric aerosol in the São Paulo Metropolitan Area, aerosols were sampled during the winter months of August 1999 and August 2000. A micro-orifice uniform deposit impactor (MOUDI) was used. Samples were submitted to gravimetry, as well as to proton-induced X-ray emission (PIXE), carbon (black and organic) and ion-chromatography analysis. These analyses supplied information about mass concentrations and physicochemical properties of the particles. Due to the higher humidity, which can increase soluble particles diameters, and reduced atmospheric stagnation seen in 2000, average PM₁₀ concentrations were higher (105 μg m⁻³) in the winter of 1999 than in the winter of 2000 (60 μg m⁻³). The PIXE analysis revealed metals and metal compounds, soil-derived elements, Si-rich particles, sulfates, carbonates, chlorides and other anthropogenic air-borne particles, supposing molecules in their usual composition. Mass balance for PM_2.5 revealed significant participation of organic and black carbon, probably resulting from diesel burning by the heavy-duty fleet.     

Size distribution and diurnal characteristics of particle-bound metals in source and receptor sites of the Los Angeles Basin

Measurement of daily size-fractionated ambient PM₁₀ mass, metals, inorganic ions (nitrate and sulfate) and elemental and organic carbon were conducted at source (Downey) and receptor (Riverside) sites within the Los Angeles Basin. In addition to 24-h concentration measurements, the diurnal patterns of the trace element and metal content of fine (0–2.5 μm) and coarse (2.5–10 μm) PM were studied by determining coarse and fine PM metal concentrations during four time intervals of the day.The main source of crustal metals (e.g., Al, Si, K, Ca, Fe and Ti) can be attributed to the re-suspension of dust at both source and receptor sites. All the crustals are predominantly present in supermicron particles. At Downey, potentially toxic metals (e.g., Pb, Sn, Ni, Cr, V, and Ba) are predominantly partitioned (70–85%, by mass) in the submicron particles. Pb, Sn and Ba have been traced to vehicular emissions from nearby freeways, whereas Ni and Cr have been attributed to emissions from powerplants and oil refineries upwind in Long Beach. Riverside, adjacent to Southern California deserts, exhibits coarser distributions for almost all particle-bound metals as compared to Downey. Fine PM metal concentrations in Riverside seem to be a combination of few local emissions and those transported from urban Los Angeles. The majority of metals associated with fine particles are in much lower concentrations at Riverside compared to Downey. Diurnal patterns of metals are different in coarse and fine PM modes in each location. Coarse PM metal concentration trends are governed by variations in the wind speeds in each location, whereas the diurnal trends in the fine PM metal concentrations are found to be a function both of the prevailing meteorological conditions and their upwind sources.     

Physicochemical characterisation of diesel exhaust particles: Factors for assessing biological activity

A range of microscopy and analytical techniques have been used to investigate the physicochemical properties of diluted DEP that may be important in determining its biological activity. Transmission electron microscopy demonstrated four basic categories of particle morphology: (1) “spherulites” [individual particles]; (2) “chains” or “clusters” of spherulites; (3) “spherules”, [large bodies of spherulites]; (4) “flake-like bodies”. Image analysis of TEM photomicrographs determined empirical morphological parameters (30 nm mean spherulite diameter, aspect ratio 1.5, mean particle area 0.078 μm, equivalent spherical diameter 0.23 μm, roundness 2.76) and derived parameters (0.313 μm² surface area, 3.7 μm² pg surface area per mass and 0.042 μm³ volume) of DEP. Distributions of the particle sizes by number showed 10.1% were ultrafine (<0.1 μm), 89.5% fine (0.1–2.0 μm), 0.4% coarse (>2.5 μm), but distributions based on a mass value were different (0.01% ultrafine; 52.6% fine, 47.4% coarse). In contrast, impacted DEP contained 60.87% ultrafine, 39.13% fine and 0% coarse particles by number. Field emission scanning electron microscopy of spherulites revealed smooth surfaces and flocculated spherules with large surface areas. Electron probe X-ray micro-analysis demonstrated the presence of C, O, Na, Mg, K, Al, Si, P, S, Cl, Ca along with a range of metals (Ti, Mn, Fe, Zn, Cr), that were heterogeneous in distribution. Inductively coupled plasma mass and atomic emission spectrometry identified Mg, P, Ca, Cr, Mn, Zn, Sr, Mo, Ba, Na, Fe, S, and Si as the mobile sorbed metals readily removed during sonication in water from DEP suspensions. X-ray Diffraction confirmed previous observations of the presence of nanometer sized crystallites of disordered graphite. Comparison of microscopy and analytical results between sonicated and impacted DEP revealed a physicochemical difference that must be taken into account in any toxicological investigations.     

Sources and chemical composition of atmospheric fine and coarse particles in the Helsinki area

During April 1996–June 1997 size-segregated atmospheric aerosol particles were collected at an urban and a rural site in the Helsinki area by utilising virtual impactors (VI) and Berner low-pressure impactors (BLPI). In addition, VI samples were collected at a semi-urban site during October 1996–May 1997. The average PM_2.3 (fine particle) concentrations at the urban and rural sites were 11.8 and 8.4 μg/m³, and the PM_2.3−15 (coarse particle) concentrations were 12.8 and about 5 μg/m³, respectively. The difference in fine particle mass concentrations suggests that on average, more than one third of the fine mass at the urban site is of local origin. Evaporation of fine particle nitrate from the VI Teflon filters during sampling varied similarly at the three sites, the average evaporation being about 50–60%.The average fine particle concentrations of the chemical components (25 elements and 13 ions) appeared to be fairly similar at the three sites for most components, which suggests that despite the long-range transport, the local emissions of these components were relatively evenly distributed in the Helsinki area. Exceptions were the average fine particles Ba, Fe, Sb and V concentrations that were clearly highest at the urban site pointing to traffic (Ba, Fe, Sb) and to combustion of heavy fuel oil (V) as the likely local sources. The average coarse particle concentrations for most components were highest at the urban site and lowest at the rural site.Average chemical composition of fine particles was fairly similar at the urban and rural sites: non-analysed fraction (mainly carbonaceous material and water) 43% and 37%, sulphate 21% and 25%, crustal matter 12% and 13%, nitrate 12% and 11%, ammonium 9% and 10% and sea-salt 2.5% and 3.2%, respectively. At the semi-urban site also, the average fine particle composition was similar. At the urban site, the year round average composition of coarse particles was dominated by crustal matter (59%) and the non-analysed components (28%, mainly carbonaceous material and water), while the other contributions were much lower: sea-salt 7%, nitrate 4% and sulphate 2%. At the rural site, the coarse samples were collected in spring and summer and the percentage was clearly lower for crustal matter (37%) and sea-salt (3%) but higher for the not-analysed fraction (51%). At the semi-urban site, the average composition of coarse particles was nearly identical to that at the urban site.Correlations between the chemical components were calculated separately for fine and coarse particles. In urban fine particles sulphate, ammonium, Tl, oxalate and PM_2.3 mass correlated with each other and originated mainly from long-range transport. The sea-salt ions Na⁺, Cl⁻ and Mg²⁺ formed another group and still another group was formed by the organic anions oxalate, malonate, succinate, glutarate and methane sulphonate. Ni and V correlated strongly pointing to combustion of heavy fuel oil as the likely source. In addition, some groups with lower correlations were detected. At the rural and semi-urban sites, the correlating components were rather similar to those at the urban site, although differences were also observed.     

Size-resolved,real-time measurement of water-insoluble aerosols in metropolitan Atlanta during the summer of 2004

During the month of August 2004, the size-resolved number concentration of water-insoluble aerosols (WIA) from 0.25 to 2.0 μm was measured in real-time in the urban center of Atlanta, GA. Simultaneous measurements were performed for the total aerosol size distribution from 0.1 to 2.0 μm, the elemental and organic carbon mass concentration, the aerosol absorption coefficient, and the aerosol scattering coefficient at a dry (RH=30%) humidity. The mean aerosol number concentration in the size range 0.1–2.0 μm was found to be 360±175 cm⁻³, but this quantity fluctuated significantly on time scales of less than one hour and ranged from 25 to 1400 cm⁻³ during the sample period. The mean WIA concentration (0.25–2.0 μm) was 13±7 cm⁻³ and ranged from 1 to 60 cm⁻³. The average insoluble fraction in the size range 0.25–2.0 μm was found to be 4±2.5% with a range of 0.3–38%. The WIA population was found to follow a consistent diurnal pattern throughout the month with concentration maxima concurring with peaks in vehicular traffic flow. WIA concentration also responded to changes in meteorological conditions such as boundary layer depth and precipitation events. The temporal variability of the absorption coefficient followed an identical pattern to that of WIA and ranged from below the detection limit to 55 Mm⁻¹ with a mean of 8±6 Mm⁻¹. The WIA concentration was highly correlated with both the absorption coefficient and the elemental carbon mass concentration, suggesting that WIA measurements are dominated by fresh emissions of elemental carbon. For both the total aerosol and the WIA size distributions, the maximum number concentration was observed at the smallest sizes; however the WIA size distribution also exhibited a peak at 0.45 μm which was not observed in the total population. Over 60% of the particles greater than 1.0 μm were observed to be insoluble in the water sampling stream used by this instrumentation. Due to the refractive properties of black carbon, it is highly unlikely that these particles could be composed of elemental carbon, suggesting a crustal source for super-micron WIA.     

On the relation between the size and chemical composition of aerosol particles and their optical properties

The light scattering and absorption coefficients of fine atmospheric aerosol particles were recorded in Hungary under rural conditions in 1998–1999 by an integrating nephelometer and particle soot absorption photometer, respectively. In some cases optical properties in the fine size range were compared to those in the coarse particles. Results obtained indicate, as expected, that fine particles control the scattering and absorption caused by the aerosol. In 1999 the size distribution of aerosol particles was also monitored by means of an electric low pressure impactor (ELPI). This makes it possible the study of the relationship between the number, surface and mass concentration in the size range of 0.1–1.0 μm and the optical characteristics by also considering the chemical composition of the particles.