Genetic Regression as a Calibration Technique for Solid-Phase Extraction of Dithizone-Metal Chelates

Abstract

The application of genetic regression (GR) to reflectance spectra of a solid-phase colorimetric extraction for the determination of Hg(II) is demonstrated. GR is a technique that combines wavelengths that optimize linear regression using a genetic algorithm. Solid polystyrene (PS) beads with a molecular weight distribution of 125,000–250,000 were impregnated with Hg(II)-dithizonate, filtered, allowed to dry, and packed in quartz cuvettes, and their reflectance spectra were collected. Quantitation with the use of normal calibration methods could not be achieved because of baseline fluctuations which originate in sample packing, presentation to the spectrometer, and particle size. Results indicate that GR can easily compensate for these baseline fluctuations and provide excellent calibrations.