Feasibility of the determination of polycyclic aromatic hydrocarbons in edible oils via unfolded partial least-squares/residual bilinearization and parallel factor analysis of fluorescence excitation emission matrices

ALARCÓN, F.; BÁEZ, M. E.; BRAVO, M.; RICHTER, P.; ESCANDAR, G. M.; OLIVIERI, A. C.; FUENTES, E.

TALANTA

ELSEVIER SCIENCE BV

Lugar: Amsterdam; Año: 2013 vol. 103 p. 361 - 370

0039-9140

The possibility of simultaneous determining seven concerned heavy polycyclic aromatic hydrocarbons (PAHs) in extra virgin olive and sunflower oils was examined by using unfolded partial least-squares with residual bilinearization (U-PLS/RBL) and parallel factor analysis (PARAFAC), both applied to fluorescence excitation emission matrices. The compounds studied were benzo[a]anthracene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[a]pyrene, indeno[1,2,3-c,d]-pyrene, dibenz[a,h]anthracene and benzo[g,h,i]perylene. The analysis was carried out by fluorescence spectroscopy after microwave assisted liquid-liquid extraction and solid phase extraction on silica. The complexity of the matrix, containing native compounds presenting spectral overlapping with the PAHs and particularly being able to produce inner filter effect, make it necessary the previous sample treatment for a selective detection of PAHs. The U-PLS/RBL algorithm showed the best performance in resolving the heavy PAH mixture in the presence of both the highly complex oil matrix and other unexpected PAHs. The obtained limit of detection with the proposed method ranged from 0.07 to 2 µg kg–1. The U-PLS/RBL predicted concentrations were satisfactorily compared with those obtained by high performance liquid chromatography with fluorescence detection.