Optimization of multivariate regression methods for quantification of sulphamethoxazole and trimethoprim in pharmaceutical formulations

Abstract

A partial least-squares calibration (PLS) procedure in combination with infrared spectroscopy has been developed for simultaneous determination of sulphamethoxazole (SMZ) and trimethoprim (TMP) in powder raw material mixtures used for manufacturing commercial pharmaceutical products. Multivariate calibration modeling procedures using interval partial least squares (iPLS) were applied to select a spectral range that provided the lowest prediction error in comparison to the full-spectrum model. The experimental matrix was constructed using 52 synthetic samples and 13 commercial samples. The considered concentration ranges were 400 to 900 mg g-1 SMZ and 80 to 240 mg g-1 TMP. Spectral data were recorded between 650 and 4000 cm-1 with a 4 cm-1 resolution by Fourier transform infrared spectroscopy (FTIR) coupled with attenuated total reflection (ATR) accessory. The proposed procedure was compared with conventional procedure by high performance liquid chromatography (HPLC) using commercial samples containing SMZ and TMP. Results shown that PLS regression model combined to ATR-FTIR is a relatively simple, rapid and accurate procedure that could be applied to the simultaneous determination of SMZ and TMP in routine quality control of powder mixtures. A root mean square error of prediction (RMSEP) of 26.97 mg g-1 for SMZ and 12.23 mg g-1 for TMP was obtained after selection of better interval by iPLS. Key-words: optimization attenuated total reflection, iPLS, pharmaceutical formulations,chemometrics.