Convolution of emission derivative ratio curves of closely related fluorescent reaction products using discrete fourier functions and non-parametric linear regression method.

A spectrofluorimetric method was used for the estimation of closely related fluorescent reaction products, Fluoxetine and Olanzapine, in their mixture after derivatization of both drugs using 4-chloro-7- nitrobenzo - 2 -oxa-1,3 - diazole (NBD-Cl) in borate buffered medium (pH 9.5) to form highly fluorescent products. The method based on the use of first and second derivative ratio of the emission data along with their convolution using 8-points sin x i or cos x i polynomials (discrete Fourier functions). The proposed method facilitates their simultaneous determination despite the presence of a minor component (Olanzapine) and strong overlapped spectra of the two NBD-Cl fluorescent products of fluoxetine and olanzapine. The accurate and precise estimation of the minor component was achieved after the convolution of the derivative ratio curves. Moreover, the obtained data were subjected to non-parametric linear regression analysis (Theil's method). The work combines the advantages of convolution of derivative ratio curves using discrete Fourier functions together with the reliability and efficacy of the non-parametric analysis of data.