Impact of Regulations of Traffic Emissions on PAH Level in the Air

2 Measuring programme, locations, sampling and analysis

2.1 Measuring programme and locations
2.2 Sampling and analysis

2.1 Measuring programme and locations

Street station

The programme covered measurements at a busy street in Copenhagen (H.C. Andersens Boulevard). This allows a comparison of the measurements of PAH and other mutagens with the results obtained for the same site in 1986, 1987, 1992 and 1993 (Nielsen, 1989 and 1996, Nielsen et al. 1995b and c and 1996, Ostenfeldt 1989). In addition, NO, gas NOy, SO₂, CO, ozone, soot, particulate matter and particulate inorganic elements have been performed for many years on a daily basis on the location (HLU, 1993, Jensen et al. 1993, Municipality 1990). The traffic intensity in the street is in the magnitude of 60,000 cars each day, and the streets in the neighbourhood have also high traffic intensities. The street is situated north-south. The measuring station was at the eastern side of the street. The summer amusement park, Tivoli, is situated on the western side of the street. The station has for many years been applied in Communality and regional air quality programmes (HLU, 1993, Jensen et al., 1993, Municipality, 1990). Most of the samples for PAH analyses and mutagenicity testing were collected in the winter (40%) and spring (45%) as the major part of the previous measurements were performed in the months January-March.

Meteorological conditions

The average meteorological conditions for the 1996 samples were comparable with those for the 1992-1993 samples. Thus, the mean temperature and wind speed outside the city was 3.5 ± 2.9 °C and 5.9 ± 1.3 m/s, respectively, for the 1996 samples compared to 2.7 ± 0.6 °C and 5.5 ± 0.6 m/s, respectively, for the 1992 - 1993 samples. Despite the average temperature was almost the same the temperature variation for the 1996 samples (range: -6.4 to 21.7 °C) was larger than for the 1992 - 1993 samples (range: -4.4 to 7.5 °C). A few of the samples in 1996 were collected in the summer and autumn. The wind direction distribution for the 1996 samples was different from the average conditions as almost the same number of samples was collected from each wind sector. In contrast to the 1996 samples the wind direction distribution for the 1992 - 1993 samples was typical for Danish conditions (Larsen and Jensen, 1983) having the dominant wind direction from west (51 %), followed by winds from south (22 %), north (16 %) and east (10 %).

Local wind directions

The local wind directions at H. C. Andersens Boulevard were determined by the topography of the city and the dominant wind directions were from south and north. The roughness of the surfaces caused the local wind speeds to be 3 - 4 times lower than that measured outside the city.

2.2 Sampling and analysis

PAC sampling

24-hour samples of airborne particulate matter were collected using conventional Hi-Vol samplers with glass fibre filters. The sampling volumes were typical about 2000 m³. The filters were stored in a freezer (-18 °C) until the analysis.

Extraction

The filters were cut into small pieces and extracted ultrasonically, 30 min. each time, with dichloromethane (p.a. Merck) (2 times) and finally with acetone (p.a. Merck). The samples were protected against light both in this step and the following ones in order to avoid photolysis of the PAC. The combined extracts were divided in two equal parts. One part was concentrated to 5 ml and applied to the mutagenicity tests (see later).

Cleaning up

The other part was added known amounts of d8-dibenzothiophene, d12-triphenylene, d12-perylene and d12-coronene (internal standards) and concentrated to 1 ml. Two ml of cyclohexane (p.a. Merck) was added and the samples were concentrated to 1.0 ml. The PAH fraction of the solution was isolated by means of liquid-liquid extraction with a cyclohexane-dimethylformamide-water system (Nielsen et al., 1986) as follows:

The PAC including PAH was isolated from the aliphatic hydrocarbons by extraction of the cyclohexane with 3 times 1.0 ml of a mixture of 90% dimethylformamide (p.a. Fluka) and 10% water (MilliQ). The three dimethylformamide-water phases were collected and 2.4 ml water was added. Thus the ratio of dimethylformamide and water was 1:1 diminishing the solubility of PAH and medium polar PAC, such as oxy-PAC, but not the solubility of polar organic compounds. The PAH and oxy-PAC were extracted from the dimethylformamide-water mixture by 3 aliquots of 3.0 ml cyclohexane. The combined cyclohexane phases were extracted with 2.0 ml water in order to remove traces of dimethylformamide. After this the cyclohexane phase was dried with sodium sulphate to remove traces of water and concentrated to about 1 ml.

Recovery

The recoveries were determined to be 85-100% for a range of representative PAC, including the internal standards.

GC-MS analysis

The PAH samples were analysed by capillary gas chromatography (Varian STAR 3400 CX) using temperature programmable splitless injection, a fused silica RTX5-MS column (Restek) and ion trap mass spectrometric detection (Varian Saturn 4D).

Mutagenicity test

One ml of the 5 ml dichloromethane extract (see above under extraction) was used for gravimetric analyses. The remaining 4 ml was evaporated to almost dryness under a gentle stream of nitrogen. 3 ml of dimethyl sulfoxide (DMSO) was then added followed by evaporation of the remaining dichloromethane by means of nitrogen. The DMSO dissolved extracts were tested for mutagenic activity in the Salmonella/mammalian microsome assay. The assays were carried out by the standard plate incorporation method described by Maron and Ames (1983) using the strains TA98 and TA98NR. The tests were performed with and without a rat liver homogenate (S9-mix) with the TA98 strain and without S9-mix with the TA98NR strain. The S9-mix (1 mg protein/plate) was prepared from Arochlor 1254 induced male Wistar rat liver. The extracts of the air particulates were tested at the following concentrations: 1) undiluted (corresponding to 10-18 m³ air pr. plate, 2) 3 times diluted, 3) 9 times diluted, 4) 27 times diluted and 5) pure DMSO (negative control). Each concentration was tested in triplicate except the undiluted extract, where only one plate was used, due to the limited test material. All extracts were tested at least twice at two independent days. The number of his+ revertants were scored manually after 48 hours incubation at 37 °C. The mutagenic activities (revertants/m³) of all extracts were calculated by linear regression analyses of the linear part of the dose response curve. Positive controls were made applying 2-aminoanthracene (TA98 with S9), 2-nitrofluorene (TA98 without S9) and 1,8-dinitropyrene (TA98NR).

N-PAC

A few samples were analysed for N-PAC and nitro-N-PAC and the basic extracts of the samples were tested for mutagenicity in order to test whether the outstanding observations of Sera et al. (1994) also could be extrapolated to Danish conditions. Sera et al. (1994) has detected and identified strong potent mutagenic nitro-N-PAC in atmospheric samples and samples from diesel exhaust in Japan. Twenty-five % of the collected dichloromethane-acetone extracts were added known amounts of d7-quinoline, d9-acridine and 10-azabenzo(a)pyrene and concentrated to 1.0 ml. Two ml toluene (p.a. Merck) was added and the mixture was concentrated to 2.0 ml. The basic N-PAC in the toluene solution was extracted with 2x2.0 ml 8.25 M phosphoric acid (p.a. Merck) (Nielsen et al., 1986). The two phosphoric acid phases was combined and adjusted to a pH of about 14 with ca. 9 ml of 11 M potassium hydroxide (p.a. Merck) in an ice bath. The N-PAC were extracted from the alkaline aqueous phase with 3x2.0 ml proportions of dichloromethane. The combined dichloromethane phases were dried with sodium sulphate and concentrated to 1.0 ml and analysed by GC-MS. The solutions for the mutagenicity tests of the N-PAC fraction were prepared in the same manner except that no internal standards were added.

Inorganic gases

CO was determined by means of infrared light absorption in a two-channels instrument (Fuji Electric, ZRC) each 30 min. Measurements of NO and volatile NO_y were performed by a two chamber chemiluminescence monitor (Monitor Labs, 8840) each 30 min. NO_y was converted to NO by passing the filtered gas sample through 315 °C molybdenum chips. All NO_y species should be reduced to NO (Nielsen et al., 1995a). SO₂ was determined by its fluorescence emission by irradiation of the polluted air sample with UV light (Monitor Labs, 8850). Ozone was determined by means of its UV-absorption.

Particles and elements

Particles were sampled on a cellulose nitrate/acetate membrane filter (Millipore RA) having a pore size of 1.2 mm during 24-h periods with an air flow volume of 60 m³. The exposed spot of the filter had a diameter of 40 mm. The particle size cut in the inlet depended on the wind speed in the surroundings (Barrett, 1984), but it corresponded in average to an aerodynamical diameter of around 20 mm. The soot content on the filters was determined by reflectometric measurements. The membrane filters were also used for the determinations of TSP (total suspended particulate matter) and elements with atomic number higher than 13 (Al). The amount of TSP was determined by weighing. This was performed in a climate room with a relative humidity of 52 ± 2 % and a temperature of 23 ± 0.5 °C (Miljøministeriet, 1986). The exposed filters were kept for at least 7 days in the climate room for conditioning prior to the weighing. After weighing one fourth of the filters were punched out for the analysis of the elements by means of PIXE (Proton Induced X-ray Emission spectroscopy) (Jensen et al., 1993, Johanson and Campbell, 1988). In the PIXE analysis the particle filter part was exposed to protons of high energy (2.5 MeV). The amounts of the elements were determined by means of the composition and intensity of the emitted X-rays.