Analysis of polycyclic aromatic hydrocarbons by liquid chromatography/tandem mass spectrometry using atmospheric pressure chemical ionization or electrospray ionization with tropylium post-column derivatization

Polycyclic aromatic hydrocarbons (PAHs) with four to six rings are potent carcinogens. This study analyzed ten of the sixteen US EPA priority PAHs using reversed-phase liquid chromatography/tandem mass spectrometry (LC/MS/MS) in selected reaction monitoring mode with two ionization sources: positive atmospheric pressure chemical ionization (APCI+) or positive elecrtrospray ionization (ESI+) with tropylium post-column derivatization. Several factors were investigated, including mobile phases, stationary phases of columns and chromatographic temperature, to determine how optimal separation and sensitivity might be achieved. Methanol used as an organic mobile phase provided better sensitivities for most PAHs than acetonitrile, although some PAHs co-eluted. Acidic buffers did not increase analyte signals. Use of Restek Pinnacle II PAH columns (250 x 4.6 mm or 250 x 2.1 mm, 5 microm) with water/acetonitrile gradient at 27 degrees C made possible a good separation of the ten analytes. [M]+. were the best precursor ions in both APCI and ESI, although fluoranthene could not be detected in ESI mode when tropylium post-column derivatization was performed. [M-28]+ and [M-52]+ were the major product ions of PAHs after collision-induced dissociation, a result of neutral losses of C(2)H(4) and (C(2)H(2))(2), respectively. Chromatographic separation for PAH isomers was crucial because the mass spectra were so similar that even MS/MS could not distinguish them from each other. The recoveries of sample preparations of PAHs spiked onto air-sampling filters ranged between 77.5 and 106% with relative standard deviations between 1.1 and 15.9%. This method was validated by analyzing NIST SRM 1649a (urban dust), producing results comparable with the certified PAH concentrations. The detection limits using APCI and ESI interfaces, defined as three times the noise levels, ranged between 0.23 and 0.83 ng and between 0.16 and 0.84 ng of on-column injection, respectively.