Infrared spectroscopy applied to the drugs quality control. Part I: multivariate regression for the quantification of potassium diclofenac

Abstract

In this work the alternative method for quantification of potassium diclofenac was studied, using diffuse reflectance infrared Fourier transform spectrocopy (DRIFTS), together with partial least squares regression (PLS). Two spectra of 20 samples with different concentrations (0.15-0.50 mg/g) of potassium diclofenac in starch had been colleted in Nicolet Magna 550 spectrophotometer, with resolution of the 4 cm-1 and 32 scans. For multivariate regression 14 samples had been used for calibration set and 6 samples had been used for validation set. For the PLS modeling, the spectral information (3715-3088 cm-1 and 1634-1191 cm-1) after multiplicative scatter correction (MSC) and autoscaling preprocess, had been processed in Pirouette® 2.7. The best models with better coefficient regression (R²), these were improved using root mean square error of validation (RMSEV). The new technique propose using DRIFTS/PLS shows an excellent choices for the quality control of the productive process of factories and drugstores that produce or handle these materials in large scale, presenting low time of analysis, no sample destruction and no wastes production.