Matrix isolation analysis of polycyclic aromatic hydrocarbons : Part I: Laser-induced fluorescence spectrometry, Part II: Gas chromatography-matrix isolation fluorescence spectrometry

The utility of laser-induced Shpol'skii fluorometry and energy-selection spectroscopy (ESS) for the characterization of matrix-isolated mixtures of alkylated and halogenated PAHs was evaluated. ESS was generally inoperative in nitrogen matrices, presumably due to strong phonon interactions with the analyte molecules. ESS was observed for PAHs in argon matrices, but it could not be used to resolve mixtures of alkylated isomers.

Laser-induced Shpol'skii fluorometry yielded quasilinear spectra for most PAHs, including those with large alkyl substituents (e.g. the t-butyl group), and those with slightly polar substituents such as halogens and methoxyl groups. Selective excitation of the quasilinear absorption bands of a particular isomer often peinnitted the unambiguous identification of that component even in the presence of 1000-fold excess quantities of other isomers.

Laser-induced matrix isolation fluorometry (LIE) was found to exhibit a relative standard deviation of 2.5-8.5 %. An analytical calibration curve for 7,10-DM-BaP was linear over five decades of concentration from a detection limit of approximately 10 pg to at least 1 pg. Quantitation of 7,10-DM-BaP in a synthetic mixture of PAHs demonstrated that radiative and non-radiative energy-transfer effects were inoperative in matrix isolated samples, even in the presence of 8000-fold excess quantities of potential interferents. Neither LIF nor ESS was useful for analyzing polar PAH derivatives. The use of conventional, Shpol'skii, and fluorocarbon matrices was ineffective presumably due to strong electron-phonon interactions and (for n-alkane matrices) hydrogen bonding.

In a related project, a gas chromatography-matrix isolation fluorescence (GC MIF) interface was used to analyze those PAH mixtures requiring separation prior to fluorometric analysis. Characterization of a mixture of alkylated and halogenated anthracene derivatives illustrated the selectivity of GC MIF.

Analytical calibration curves were linear over three orders of magnitude from a blank limited detection limit of approximately 10 ng to at least 1 pg. GC MIF was observed to have a relative standard deviation of 10.7 %. Quantitation of pyrene in three NBS standard reference materials, including a shale oil and an urban dust sample, resulted in errors of 2.0-8.7 %.