Heterogeneous ozonation reactions of PAHs and fatty acid methyl esters in biodiesel particulate matter

Numerous studies have examined the oxidation of PAHs found in diesel particulate matter (PM) by ozone, but no studies have investigated the ozone oxidation of biodiesel exhaust PM. Fatty acid methyl esters (FAMEs), found in high abundance in biodiesel PM, can potentially alter the kinetics of the reactions between atmospheric oxidants such as ozone and particle-phase PAHs. In this study, the heterogeneous reactivity of 16 EPA PAHs upon 24 h exposure to 0.4 ppm ozone in the presence (PAH + FAMES) and absence (PAH-only) of FAMEs was investigated at room temperature and 50% relative humidity. The ozone-reactivity of the PAHs detected in 20% biodiesel (B20) exhaust PM was also investigated. In the absence of FAMEs, the pseudo-first order ozone reaction rate constant, kO3,of PAHs varied from 0.086 ± 0.030 hr−1 (chrysene) to 0.184 ± 0.078 hr−1 (anthracene). In the presence of FAMEs, kO3 of the PAHs varied between 0.013 ± 0.012 hr−1 (benzo[b]fluoranthene) and 0.168 ± 0.028 hr−1 (benzo[a]pyrene), and with the exception of benzo[a]pyrene, the kO3of PAHs were 1.2–8 times lower compared to those obtained during the PAH-only ozone exposure. Only one PAH, benzo[a]pyrene (BaP), did not show a significant change in kO3 with addition of FAMEs. Phenanthrene, fluoranthene, and pyrene, the only PAHs detected in the B20 PM, had kO3 values about 4 times lower in B20 PM than those obtained when spiked PAHs-only were exposed to ozone. The kO3 values of phenanthrene and fluoranthene in the B20 PM were 2 times higher than rates obtained when the PAH mix was exposed to ozone in the presence of the FAMEs. In contrast, pyrene's kO3 in the B20 PM was about 2 times lower than that obtained for the PAH + FAMEs exposure. Observed differences in PAH behavior demonstrate individual PAH heterogeneous reactivity with gas-phase ozone is sensitive to PAH (vapor pressure, solubility/sorption to matrix components, chemical reactivity) as well as substrate properties (PAH and O3 diffusivity in the matrix that may evolve with reaction progress). Saturated FAMEs were not reactive with ozone (kO3 range = 0.004 ± 0.003 to 0.012 ± 0.026 hr−1), but compared to PAHs, up to two times higher kO3 was measured for the unsaturated FAMEs (range 0.087 ± 0.015 to 0.329 ± 0.023 hr−1) during PAH + FAMEs exposures. These changes in substrate composition during atmospheric aging would be expected to affect PAH diffusivity and therefore heterogeneous reactivity over time. The factor of 1.2–8 decreased heterogeneous reactivity of PAHs in the presence of the FAMEs mix and the B20 PM matrix suggests that the presence of FAMEs in the diesel fuel supply may lead to increased PAH atmospheric lifetimes and longer range PAH transport. Predictive methods to quantify changes in PAH reactivity with gas-phase oxidants as a function of substrate composition and characteristics (viscosity, polarity, degree of unsaturation) are needed as biodiesel is increasingly present in our diesel engine fuel supply from a variety of feedstocks at different blend ratios.