Evaluation of a denuder-MOUDI-PUF sampling system to measure the size distribution of semi-volatile polycyclic aromatic hydrocarbons in the atmosphere

This study describes a field comparison conducted between 2 methods employing different MOUDI impactor configurations to evaluate their performance in sampling and measurement of the size distribution of 15 priority pollutant polycyclic aromatic hydrocarbons (PAHs). Samples were collected during 24 h periods approximately every 7th day, beginning at 8:00 AM, in 2 different sites of the Los Angeles Basin. One site was near Central Los Angeles in an area impacted by high vehicular traffic, whereas the other site was located about 60 km downwind of central Los Angeles (receptor site). Particle samples from about 43 m(3) of air were collected using collocated MOUDI impactors and classified in 3 aerodynamic diameter size intervals: 0-0.18 mum (ultrafine mode I), 0-18-2.5 mum (accumulation mode II), and 2.5-10 mum (coarse mode III). One MOUDI operated in the conventional mode, the other with a vapor trapping system that included an XAD-4 coated annular denuder placed upstream of the impactor and a polyurethane foam plug (PUF) placed in series behind the impactor. PAHs were separated and quantified by HPLC with fluorescence detection optimized for the highest sensitivity. The results showed that for both sites, using either sampling system, the size distributions obtained are similar for the less volatile PAHs (log [pdegreesL] less than or equal to - 3.2), but different for the more volatile PAHs (log [pdegreesL] greater than or equal to -2.06). In the central Los Angeles site, the largest PAH fraction was found in the 0-0.18 mum (mode I) size range, typical of primary emissions. At the downwind location, the largest fraction was in the 0.18-2.5 mum (mode II) size range, consistent with an "aged" aerosol. At both sites, albeit not statistically significant, the mean regular to denuded MOUDI mass ratios were 33-36% and 11-19% higher, respectively, for the more volatile and the less volatile PAH groups. Sampling with the regular MOUDI configuration is simpler and thus recommended for measurement of the size distribution of PAHs in either group.