LAUSR

Abstract This study investigated the gas–particle partition and spatial characteristics of polycyclic aromatic hydrocarbons (PAHs) in the ambient air of a prototype coking plant. High-volume samplers with glass fiber filters (GFFs) and polyurethane foams (PUFs) were used to collect PAHs in both gaseous and particulate phases. The EPA's 16 priority-controlled PAHs were measured using gas chromatography equipped with a mass-selective detector (GC-MS). The total PAH concentrations ranged from 1.32 μg m−3 to 9.41 μg m−3, with an average of 4.20 μg m−3. PAH abundance at the various sites of the coking plant was determined to be in the following order: coke side > machine side > leeward site; the lowest values were at the reference site and factory boundary. The medium-molecular-weight PAHs had the highest fraction at the coke side, whereas the low-molecular-weight PAHs were predominant at other sampling sites. Considerable amount of chrysene was detected at the coke side, fluorene at the machine side, acenaphthene at the leeward site (240 m downwind of the coke oven), and naphthalene at the other three sites. The gaseous fractions of PAHs at the six sampling sites ranged from 46.3% to 85.8% of the total PAHs in the gas–particle systems; and gaseous PAHs were predominant in the ambient air of the coking plant—except for the coke side, where the PAHs were dominantly partitioned to the particles. In addition, the particulate benzo(a)pyrene equivalence (BaPeq) fraction was high in the coking plant air. The slope (mr) values ranged from −0.59 to −0.39 in the plant, indicating that the absorption mechanism, rather than the adsorption mechanism, dominated the gas–particle partitioning. The intercept (br) values carry the information on the gaseous–particulate phase distribution of semivolatile organic compounds when the mr values are similar. The organic matter fraction in particles, particulate phase fraction of PAHs, and toxicity of particulate matter were found to increase with an increase in br values at similar mr values. The predominant mechanism was absorption for dibenz(a,h)anthracene, benzo(ghi)perylene, and indeno(1,2,3-cd)pyrene, but adsorption for benzo(b)fluoranthene, benzo(a)pyrene, and benzo(k)fluoranthene at the coke and machine sides. By contrast, for most high-molecular-weight PAHs, absorption into the organic matter was the predominant mechanism at other sites. The gas–particle partitioning mechanism of individual PAHs was closely related to the distance from the coke oven, rather than the orientation to the oven. These results may be of great importance in understanding the pollutant process and atmospheric transport of PAHs and may provide guidance to decision makers regarding air pollution control.