Characterization of PM 2.5-bound polycyclic aromatic hydrocarbons in Atlanta

Abstract

Twenty-four hour PM 2.5 samples from a rural site, an urban site, and a suburban site (next to a major highway) in the metropolitan Atlanta area in December 2003 and June 2004 were analyzed for 19 polycyclic aromatic hydrocarbons (PAH). Extraction of the air samples was conducted using an accelerated solvent extraction method followed by isotope dilution gas chromatography/mass spectrometry determination. Distinct seasonal variations were observed in total PAH concentration (i.e. significantly higher concentrations in December than in June). Mean concentrations for total particulate PAHs in December were 3.16, 4.13, and 3.40 ng m −3 for the urban, suburban and rural sites, respectively, compared with 0.60, 0.74, and 0.24 ng m −3 in June. Overall, the suburban site, which is impacted by a nearby major highway, had higher PAH concentration than did the urban site. Total PAH concentrations were found to be well correlated with PM 2.5, organic carbon (OC), and elemental carbon (EC) in both months ( r 2 = 0.36–0.78, p < 0.05), although the slopes from the two months were different. PAHs represented on average 0.006% of total PM 2.5 mass and 0.017% of OC in June, compared with 0.033% of total PM 2.5 and 0.14% of OC in December. Total PAH concentrations were also correlated with potassium ion ( r 2 = 0.39, p = 0.014) in December, but not in June, suggesting that in winter biomass burning can potentially be an important source for particulate PAH. Retene was found at a higher median air concentration at the rural site than at the urban and suburban sites—unlike the rest of the PAHs, which were found at lower levels at the rural site. Retene also had a larger seasonal difference and had the weakest correlation with the rest of the PAHs measured, suggesting that retene, in particular, might be associated with biomass burning.