The use of the Box-Behnken experimental design in the optimisation and robustness testing of a capillary electrophoresis method for the analysis of ethambutol hydrochloride in a pharmaceutical formulation.

Box-Behnken experimental designs do not appear to be extensively used in optimisation of analytical methods using capillary electrophoresis (CE). This paper describes the use of the Box-Behnken experimental design to optimise the factors affecting the separation of ethambutol hydrochloride (EB), its impurity 2-amino-1-butanol and the internal standard (phenylephrine hydrochloride) in a CE method for a pharmaceutical tablet assay. The three factors studied simultaneously were: buffer pH, buffer concentration and applied electric field, each at three levels. The method was optimised with respect to three responses: resolution between peaks, theoretical plate count and the migration time of the EB peak. A statistical programme, which applies a multiple response optimisation algorithm, was used to calculate and optimise the three responses simultaneously. The optimum conditions were established to be 58.0 mM sodium borate buffer at pH 9.50 and an applied electric field of 412 V/cm. The robustness of the method was also determined and confirmed using a second Box-Behnken design, as part of the validation exercise. System suitability values for the method were derived from the regression surface analysis. The CE method for a pharmaceutical tablet formulation was further validated according to current regulatory requirements, with respect to linearity and range, precision, specificity, accuracy and limit of quantitation. The optimised method gives a fast and efficient separation under 4 min, with complete resolution between the three peaks, and represents an improvement over the existing USP method. It can be concluded that the Box-Behnken experimental design provides a suitable means of optimising and testing the robustness of a CE pharmaceutical method.

Full and fractionated experimental designs for robustness testing in the high-performance liquid chromatographic analysis of codeine phosphate, pseudoephedrine hydrochloride and chlorpheniramine maleate in a pharmaceutical preparation.

This paper describes the testing of a saturated factorial design using a full factorial design. Saturated factorial designs are often used to test the robustness of high-performance liquid chromatography (HPLC) methods, however they are based on several assumptions. A full factorial design relies on fewer assumptions and hence could be used to evaluate the effectiveness of the saturated design. Both designs were used to test a gradient HPLC method for the assay of codeine phosphate, pseudoephedrine hydrochloride and chlorpheniramine maleate. Six HPLC conditions, including wavelength, mobile phase pH and ion pairing reagent concentration were tested using the saturated design. Three of these factors were selected for full evaluation using a full factorial design. The results showed that the main effects calculated by each design were comparable. However, the saturated design showed higher standard errors, probably due to the effects of changing several more factors. One interaction effect was indicated as a confounding effect by the saturated design and this was confirmed by the calculation of the same interaction effect using the full design. Overall the method was shown to be robust under the variety of HPLC conditions tested.