Authors: Marcellos, L.F.; Faria, A.F.; De Souza, M.V.N.; Almeida, M.R.; Sabin, G.P.; Poppi, R.J.; Oliveira, M.A.L.
Source: Central European Journal of Chemistry, v. 10, p. 1808-1816, 2012.
Publisher: De Gruyter
Abstract
The development and optimization of a novel UV spectrophotometric methodology was proposed for simultaneous analysis of ethambutol (ETB), isoniazid (ISO), rifampicin (RIF) and pyrazinamide (PYR), using multivariate calibration based on the partial least squares method (PLS). The methodology was successfully applied for analysis of four-drug fixed dose combination (4-FDC) tablets used for tuberculosis treatment. A 34 Box-Behnken design, with triplicate in central point, was used for sample preparation in the calibration step. In the present case, nine latent variables were chosen for the model development that presented the smallest RMSECV and explain 98.76% of data variance in Y block (concentrations of ETB ISO, RIF and PYR) and 99.93% of data variance in X block (spectral data). PLS models for ETB, ISO, RIF and PYR presented RMSEP and R2 values of 0.23 mg L−1 and 0.971; 0.14 mg L−1 and 0.731; 0.11 mg L−1 and 0.990 and 0.57 mg L−1 and 0.972, respectively. A validation step was performed based on the comparison between the UV spectrophotometric proposed methodology and capillary zone electrophoresis (CZE) in 4-FDC real samples and no significant difference was found between two methodologies at 95% of confidence levelKeywords: 4-FDC; UV spectrophotometry; Chemometrics; Capillary electrophoresis Document Type: Research Article DOI: http://dx.doi.org/10.2478/s11532-012-0102-6 Publication date: 21 de Setembro de 2012
