2011_04_06_Handling of Rayleigh and Raman scatter for PARAFAC modeling of fluoresce



ESEL Paper Review_20110404 by Aamir Alaud-din

1.    Title and Author

Title: 2011_04_04_Handling of Rayleigh and Raman scatter for PARAFAC modeling of fluorescence data using interpolation
Authors: Morteza Bahram1, Rasmus Bro2*, Colin Stedmon3, and Abbas Afkhami1

1Department of Chemistry, Faculty of Sciences, Bu-Ali Sina University, Hamadan, Iran

2DChemometric Group, Food Technology, Department of Dairy & Food Science, Royal Veterinary & Agricultural University, Rolighedsvej 30, DK-1958 Frederiksberg C, Denmark

3Department of Marine Ecology, National Environmental Research Institute, P.O. Box 358, Frederiksborgvej 399, DK-400 Roskilde, Denmark

2.    Summary of Paper

Fluorescence excitation-emission matrix (EEM) is very useful in a lot of fields like food science, analytical chemistry, chemometrics, and environmental science and engineering. EEM data contains a lot of useful information that can be obtained using parallel factor analysis (PARAFAC) modeling. Problems arise in extracting information from EEM data because of Rayleigh and Raman scatters. Several methods have been proposed to handle these scatter peaks. This paper deals with these scatters using interpolation. Advantage of this technique is the simplicity, algorithm speed and enabling of software to handle missing data. Results obtained this way were compared with already published results and this technique is proved successful.

To validate the technique, three different fluorescence data sets were used. Dataset-I was a mixture of four different fluorophores. Excitation wavelength range was 200 ? 350 nm with a 5 nm interval. Emission wavelength range was 200 ? 750 nm. Dataset-II consisted of solutions from sugar industry. Seven different excitation wavelengths viz., 230, 240, 255, 290, 305, 325, and 340 nm were used. Emitted signals range was 275 ? 560 nm. Dataset-III was the sea water samples. Excitation and emission wavelength ranges with their respective intervals were 240 ? 600 nm with 5 nm intervals and 240 ? 450 nm with 5nm interval.

Results were successful, and loadings from PARAFAC models for dataset-II are shown below.

Loadings from PARAFAC modeling of dataset-II

3.    Contribution to ESEL

This paper shows that a very simple technique may sometimes be very useful to handle complex situation. Simple techniques require less time and sometimes give comparable results to advanced methods.

By: Aamir Alaud-din

첨부 (1)
191.7KB / Download 2