Validation studies of thermal extraction-GC/MS applied to source emissions aerosols. 1. Semivolatile...

In this work, we develop a novel validation approach for studying how nonvolatile aerosol matrixes of considerably different chemical composition potentially affect the thermal extraction (TE)-GC/MS quantification of a wide range of trace semivolatile organic markers. The nonvolatile matrixes of a set of source emissions aerosols are first operationally isolated by thermally clearing the aerosols of their native semivolatile organic matter. TE-GC/MS analysis is then performed in triplicate on matrixes refortified with multilevel organic compound standard suites. The spiking of empty thermal extraction tubes and blank quartz filters is introduced as experimental control and also allows for the calculation of method detection limits. For the vast majority of organic compounds fortifying the matrixes (e.g., the alkane, alkene, cycloalkane, sterane, and phthalate classes), the analytical bias observed was classified as either minor or nonexistent. Furthermore, compound recoveries were generally highly reproducible, demonstrating relative standard deviations of less than 20%. For a diesel engine exhaust sample, significant matrix effects for the six- and seven-ring polycyclic aromatic hydrocarbons (PAHs) are observed and ascribed to the high proportion of elemental carbon in the sample. Our results suggest that TE-GC/MS may underestimate inhalation exposures to PAHs (with 5 rings or more) in atmospheric aerosols replete with diesel engine exhaust (e.g., near roadways or in polluted urban air). Due to its stability and representativeness, the use of a thermally cleared particulate matter matrix for validation purposes is probably expandable to additional sample pretreatment and instrumental techniques also being applied to quantify organic molecular markers in source and atmospheric aerosols.