Temperature-dependent emission concentrations of polycyclic aromatic hydrocarbons in paving and built-up roofing asphalts.

OBJECTIVES To characterize temperature-dependent emissions from paving and built-up roofing asphalt (BURA) and to quantify differences in temperature-related concentrations and composition. METHODS Using headspace gas chromatography, 18 polycyclic aromatic hydrocarbon (PAH) emission concentrations were quantified over eight temperatures (120-315°C) for paving asphalt (n = 20) and Types II, III, and IV BURA (n = 5) and were summarized by geometric means (GMs) and geometric standard deviations (GSDs) at each temperature. The relationships between temperature and concentration were evaluated for PAH analytes using mixed-effects regression models. Temperature was categorized into regimes: Regime 1 (120-150°C) representing temperatures typical of paving asphalt application, Regime 2 (180-230°C) representing temperatures typical of BURA application, and Regime 3 (260-315°C) which were high temperatures outside typical application temperatures. An interaction term was used to evaluate differential effects of temperature on paving asphalt versus BURA. RESULTS In the paving regime (120-150°C), paving asphalt emission concentrations were highest for 2- and 3-ring PAHs [GM (GSD) at 150°C of 4.51 (2.07), 3.77 (1.63), 2.26 (1.53), and 1.80 (1.66) μg m(-3) for 2-methyl naphthalene, naphthalene, phenanthrene, and acenaphthene, respectively], with all the 4- and 5- to 6-ring individual PAHs mean concentrations below the detection limit, with the exception of benz[a]anthracene. In the BURA regime (180-230°C), BURA emission concentrations were highest for 2- and 3-ring PAHs [GM (GSD) at 205°C of 121.3 (1.37), 99.5 (1.31), 69.5 (1.32), and 68.1 (1.37) μg m(-3) for acenaphthene, anthracene, 2-methyl naphthalene, and phenanthrene, respectively], with lower but detectable concentrations for 4- and 5- to 6-ring PAHs. For both paving asphalt and BURA, concentrations increased log linearly with temperature. At a given temperature, the highest concentrations were observed for 2-ring PAHs with lower and decreasing concentrations observed with increased ring size. Temperature was a statistically significant (P < 0.01) predictor of concentration for each analyte. Furthermore, the categorical temperature regime variable explained a large percent of the variability in concentrations accounting for 74-92% of the total variability in PAH concentration. In both paving asphalt and BURA, the relationship between temperature and concentration was non-linear. There was a statistically significant difference between paving asphalt and BURA in the temperature-concentration relationship for the each analyte. Temperature alone did not account for differences in paving asphalt and BURA concentrations in 5- to 6-ring PAHs. CONCLUSIONS Emission concentrations are driven by temperature for both paving asphalt and BURA samples under studied experimental conditions. There are differences in paving asphalt and BURA emission concentrations that are not explained by temperature alone; concentrations were higher and consisted of larger ring (4 and 5 to 6) PAHs for BURA as compared to paving asphalt at the respective application temperature ranges.