Method development in supercritical fluid chromatography using design of experiments and design space : Application to the separation optimization of an antiepileptic drug substance and its impurities

Abstract

In the context of promoting green chemistry, an optimized method was developed using Supercritical Fluid Chromatography (SFC) associated to Design of Experiments (DoE) methodology and Response Surface Modeling (RSM). In this study, the performance of SFC was challenged by a real pharmaceutical case-study, which was applied to the optimal separation of an anti-epileptic drug substance and its impurities. The first step included a screening of different injection solvents, stationary phase chemistries (columns packed with sub-2µm particles) and mobile phase compositions. To this end, a full factorial design was carried out. The combination of CPME as diluent solvent, 1-aminoanthracene as stationary phase and a mobile phase composed of mainly CO2 and methanol + 2% of water as modifier provided the best peak shapes with acceptable values for the tailing factor (t_f), maximized peak capacity (Pc) and peak height (h). Secondly, the method was optimized by investigating the percentage of water as additive in the mobile phase composition (%H2O), the backpressure (BP), and the gradient time (t_G) over a predefined experimental domainin order to find the optimal separation. Therefore, all 17 drug compounds were eluted within 6.6 min gradient time when setting the backpressure at 120 bar and adding 2% of H2O as additive in the mobile phase. DoE was used to perform the experiments. Retention times at the peak apex (RT), at the peak start (〖RT〗_start), at the peak end (〖RT〗_end) and peak width at half height (w_0.5) were modeled for 216 conditions over the experimental domain. Based on these responses, two critical quality attributes (CQAs) were computed: the minimal resolution (Rs) and the minimal separation (S) criteria. The two approaches and their respective resulting number of coelutions predictions were compared to determine the design space (DS) providing optimal separation conditions. Excel® was used to process all data and to simulate the predicted chromatograms, easing the application of this study outcome to other case studies. Once again, SFC associated to DoE and computer-assisted optimization methodology was demonstrated to be a powerful tool and a green alternative for method development in the context of impurity profiling.