Second-order calibration is used for second-order data. Such data are produced by instruments that give a matrix of responses for a single measured standard or unknown sample. This allows determination of analyte of interest in the presence of uncalibrated sample constituents, a property known as the second-order advantage. It was discovered that a second order spectra data matrix of sunset yellow produced from the solvent components gradual change–visible absorption spectra can be expressed as the combination of two bilinear data matrices. Based on this discovery, a new method for the determination of sunset yellow in gray systems using second order calibration algorithms has been developed. The second order calibration algorithm was based on the rank annihilation factor analysis (RAFA). In the method described here, the components of the solvent were changed gradually by adding dioxan into water, the absorption spectra of sunset yellow and saffron samples in a series of water–dioxan mixed solvents with various dioxan volume fractions were recorded, and then the second order data were obtained from the solvent components gradual change–visible absorption spectra. Thus, the concentration of sunset yellow in adulterated saffron samples could be determined from the spectra matrices using second order calibration algorithms. This method is simple, convenient and dependable. The method has been successfully used to determine sunset yellow in saffron samples with satisfactory results.

Keywords