Renewal of an old European Pharmacopoeia method for Terazosin using modeling with mass spectrometric peak tracking

An older method for terazosin was reworked in order to reduce the analysis time from 90 min (2 x 45 min) to below 5 min. The method in European Pharmacopoeia (Ph.Eur.) investigates the specified impurities separately. The reason of the different methods is that the retention of two impurities is not adequate in reversed phase, not even with 100% water. Therefore ion-pair-chromatography has to be applied and since that two impurities absorb at low UV-wavelength they had to be analyzed by different method than the other specified impurities. In our new method we could improve the retention with pH elevation using a new type of stationary phases available for high pH applications. Also a detection wavelength could be selected that is appropriate for the detection and quantification of all impurities. The method development is the bottleneck of liquid chromatography even today, when more and more fast chromatographic systems are used. Expert knowledge with intelligent programs is available to reduce the time of method development and offer extra information about the robustness of the separation. Design of Experiments (DoE) for simultaneous optimization of gradient time (t(G)), temperature (T) and ternary eluent composition (t(C)) requires 12 experiments. A good alternative way to identify a certain peak in different chromatograms is the molecular mass of the compound, due to its high specificity. Liquid Chromatography Mass Spectrometry (LC-MS) is now a routine technique and increasingly available in laboratories. In our experiment for the resolution-and retention modeling the DryLab4 method development software (Version 42) was used. In recent versions of the software the use of (m/z)-MS-data is possible along the UV-peak-area-tracking technology. The modelled and measured chromatograms showed excellent correlations. The average retention time deviations were ca. 0.5 s and there was no difference between the predicted and measured R-s,R-crit values. (C) 2016 Published by Elsevier B.V.