Solid-liquid extraction room temperature phosphorimetry and pattern recognition for screening polycyclic aromatic hydrocarbons and polychlorinated biphenyls in water samples

A ARRUDA,; GOICOECHEA, HÉCTOR C; M SANTOS,; A CAMPIGLIA,; AC OLIVIERI,

ENVIRONMENTAL SCIENCE & TECHNOLOGY

American Chemical Society

Año: 2003 vol. 37 p. 1385 - 1391

0013-936X

Solid-liquid extraction room temperature phosphorimetry and pattern recognition are being proposed for screening polycyclic aromatic hydrocarbons (PAH) and polychlorinated biphenyls (PCB) in water samples. Both classes of pollutants are extracted with the same membrane using a syringe kit, and their phosphorescence spectra are recorded directly from the surface of the solid support. The entire experimental procedure takes less than 8 min per sample, and the limit of detection for 10 mL of water is at the parts-per-billion level. On the basis of the sample spectrum, pattern recognition is used to classify positive water samples as containing PAH, PCB, or both. The feasibility of this approach and its ability to avoid unnecessary chromatographic analysis of uncontaminated samples is demonstrated with several complex mixtures and the analysis of heavily contaminated “real world” samples. The recognition ability was 100%, while the prediction ability was close to perfect.-liquid extraction room temperature phosphorimetry and pattern recognition are being proposed for screening polycyclic aromatic hydrocarbons (PAH) and polychlorinated biphenyls (PCB) in water samples. Both classes of pollutants are extracted with the same membrane using a syringe kit, and their phosphorescence spectra are recorded directly from the surface of the solid support. The entire experimental procedure takes less than 8 min per sample, and the limit of detection for 10 mL of water is at the parts-per-billion level. On the basis of the sample spectrum, pattern recognition is used to classify positive water samples as containing PAH, PCB, or both. The feasibility of this approach and its ability to avoid unnecessary chromatographic analysis of uncontaminated samples is demonstrated with several complex mixtures and the analysis of heavily contaminated “real world” samples. The recognition ability was 100%, while the prediction ability was close to perfect.