Chemometrical Evaluation of Metoprolol Tartarate Enantiomers Separation Applying Conventional Achiral Chromatography

Most β-adrenergic antagonists (β-blockers) are therapeutically used as a racemic mixture and their enantiomers demonstrate different pharmacodynamic and pharmacokinetic properties [1,2]. The both enantiomers are reducing blood pressure, but the therapeutic effects of the (S)-enantiomer is about 100 times stronger than (R)-enantiomer. The increasing demands for the production of enantiomerically pure drugs have led to enantioselective separations becoming one of the most important analytical tasks. In this context, enantioselective HPLC is one of the most powerful and widely employed separation techniques, both for analytical and preparative purposes, as well as for research in pharmaceutical and biomedical analysis.

Most β-adrenergic antagonists (β-blockers) are therapeutically used as a racemic mixture and their enantiomers demonstrate different pharmacodynamic and pharmacokinetic properties [1,2]. The both enantiomers are reducing blood pressure, but the therapeutic effects of the (S)-enantiomer is about 100 times stronger than (R)-enantiomer. The increasing demands for the production of enantiomerically pure drugs have led to enantioselective separations becoming one of the most important analytical tasks. In this context, enantioselective HPLC is one of the most powerful and widely employed separation techniques, both for analytical and preparative purposes, as well as for research in pharmaceutical and biomedical analysis.
The aim of our investigation was analysis of metoprolol tartarate enantiomers using achiral stationary phase with β-cyclodextrin as chiral additive in the mobile phase. The experimental design was applied as the best way to define chromatographic behavior of enantiomers. Analysis employed ″one factor at a time″ method for preliminary study, full factorial design for screening of experiment and response surface methodology (RSM) for method optimization. A literature search showed many experimental design applications in analytical method development and validation, especially in the area of separation science. Experimental design has been used for separation optimization [10][11][12] and for validation in RP-HPLC method [13,14]. It was used for robustness testing in RP-HPLC method [15,16] and capillary electrophoresis [17]. The methodology proposed in this paper represents novel, efficient and easily attainable approach in resolving metoprolol tartarate enantiomers using conventional achiral chromatography.

Chromatographic conditions
The chromatographic system Hewlett Packard 1100 (Agilent, Technologies) consisted of a HP 1100 pump, HP 1100 UV-VIS detector and HP ChemStation integrator. Separations were performed on a Supelcosil LC 18 4.6 mm × 250 mm, 5 µm particle size column.
UV detection was performed at 275 nm. The samples were introduced through a Rheodyne injector valve with a 20 µL sample loop.

Buffer solution
Buffer solution was prepared by adding of 2 ml of TEA to 600 ml of HPLC water, pH was adjusted in the range from 2 to 6 with glacial acetic acid.

Solution for equlibration and storage of column
Solution was prepared in concentration of 3 mM of β-CD in water.

Results and Discussion
In preliminary investigations, influence of different chromatographic factors on separation of metoprolol tartarate enantiomers was analyzed. As the separation was performed employing conventional achiral chromatography, beta cyclodextrine (β-CD) was added to the mobile phase as chiral modificator. Mobile phases consisted of acetonitrile and β-CD in triethylamine/glacial acetic acid buffer in different ratios. Column was stored and conditioned with solution of 3 mM of β-CD in water. In chromarography many factors can influence separation e.g., content of organic modifier in mobile phase, pH of the mobile phase, column temperature, flow rate, concentration of solute etc. In the first step of our study, pH of the mobile phase and temperature were defined using "one factor at a time" method. pH of the mobile phase (2.5; 3.0; 4.0; 5.0 and 6.0) was changed and other factors were kept at constant level. Obtained chromatograms demonstrated the best separation of enantiomers at pH 3.0 and in following investigations it remained constant. Secondly, temperature was analyzed on two levels 30°C and 40°C. The accepted separation was at 35°C column temperature.
In the second step, for the screening of experiment, full factorial design 2 4 was chosen. Full factorial designs at two levels are mainly used for screening, that is, to determine the influence of a number of effects on a response and to eliminate those that are not significant [18]. Selected factors and their "low" (-) and "high" (+) levels are presented in Table 1. Matrix of the experiment is given in Table 2.
In experimental design for the evaluation of influence of investigated factors, on measured response, mathematical model was applied. Often form of a mathemathical model is: where y presents the estimate response, b 0 , is the average experimental response, the coefficients b 1 to b N are the estimated effects of the factors considered and the extend to which these terms affect the performance of the method is called main effect. The coefficients b 12 to b (N-1)N are called the interaction terms. We can see that the factorial design provides information about the importance of interaction between the factors [18]. The calculating coefficients of mathematical models for outputs are presented in Table 4.
The results showed that acetonitrile content and concentration of β-CD had the biggest influence on retention factors (Y 1 and Y 2 ). The flow rate and content of β-CD influenced selectivity factor (Y 3 ) and resolution (Y 4 ) the most. Concentration of metoprolol tartarate had negligible influence on analyzed outputs and in further investigations it was kept constant.
In the third step of method optimization, three factors (content of acetonitrile, content of -CD and flow rate) were analyzed in 22 experiments. Matrix of experiment for optimization is presented in Table 5. Experimental data for outputs are presented in Table 6.
On the basis of the results, outputs Y 1 and Y 4 were chosen to analyze separation. The results for others outputs gave bad coefficient of determination (≤ 0.5) and they fitted badly in the obtained model. Those results could be explained with different characters of     Table 7 for output Y 1 and in Table 8 for output Y 4 .
Coefficients of determination (R 2 ) and results for factor Fisher value (F) demonstrated good fitting of obtained results in mathematical model. Suitable three-D graphs are presented in Figure 1 for output Y 1 and in Figure 2 for output Y 4 .
Obtained three-D graphs gave information about influence of acetonitrile content and chiral modificator concentration on metoprolol tartarate enantiomers separation. The connection between influence of the factors and outputs can be presented with second order polynoms. The obtained polynoms are presented as Equation 2 and Equation 3.
where is Y 1 -is retention factor of enantiomer R (+) where is Y 4 -is resolution factor As it could be seen from the Figure 1, retention factor of the first enantiomer has a higher value for higher content of β-CD and higher content of acetonitrile. Also, strong influence of acetonitrile content is obvious and drastical decrease of retention factor was observed. On the other hand, resolution factor is under strong influence of both investigated factors. It is clear that both factors must be carefully set in order to achieve acceptable separation of enantiomers investigated.
According to the presented results the best separation of metoprolol tartarate enantiomers can be achieved with mobile phase:         Under these conditions the value of the resolution factor is 0.98. The representative chromatograms are presented in Figure 3.

Conclusion
Applying experimental design it was possible to achieve optimal separation of metoprolol tartarate enantiomers performing a relatively small number of experiments. In this paper full factorial design 2 4 was used for experiment screening. Chromatographic behavior of investigated enantiomers was affected by acetonitrile content, chiral modificator content in the mobile phase and flow rate the most, which was demonstrated by obtained linear models. After experimental screening, RSM was used for optimization of RP-HPLC method and optimal chromatographic conditions were settled. The proposed methodology represents an efficient and easily attainable approach in resolving the problem of searching for optimal HPLC chromatographic conditions via experimental design.