Free Access
Volume 28, Number 2, March 2000
New trends in material science
Page(s) 148 - 154
Section Original articles
Analusis 28, 148-154 (2000)
DOI: 10.1051/analusis:2000108

Modeling of inner filter effect in synchronous spectrofluorimetry by using partial least squares

A. Andrade-Eiroa, E. Vázquez-Blanco, P. López-Mahía, S. Muniategui-Lorenzo and D. Prada-Rodríguez

Department of Analytical Chemistry. A Coruña University. A Zapateira, E 15071, A Coruña. Spain

(Received September 29, 1999; revised January 4, 2000; accepted February 8, 2000.)

This paper deals with a very common problem on spectrofluorimetry: the inner-filter effect. This limitation has made spectrofluorimetry into a disadvantageous technique for the quantification of individual compounds in complex mixtures, however, a lot of papers have quantified compounds as fluorene in mixtures where other Polycyclic Aromatic Hydrocarbons (PAHs) absorb part of the energy emitted by fluorene without taking account this limitation. The inner-effect filter for fluorene is easily detectable in spectrofluorimetric measurements in mixtures where there are compounds, such as benzo(k)fluoranthene, indeno(1,2,3-cd)pyrene and pyrene. The application of methods, such as Multiple Linear Regression (MLR) to the quantification of fluorene in mixtures containing compounds capable to quench its signal, provides high errors in the analytical results. This point is carefully treated in our paper. The precise and accurate quantification of fluorene in presence of benzo(k)fluoranthene, indene(1,2,3-cd)pyrene and pyrene, was achieved by constant-wavelength synchronous fluorimetry in combination with Partial Least Squares (PLS) calibration.

Key words: Partial least squares -- polycyclic aromatic hydrocarbons -- multiple linear regression - chemiometric.

© EDP Sciences, Wiley-VCH 2000