OPTIMIZATION OF PARAMETERS OF A GAS-CHROMATOGRAPHIC METHOD

Zsuzsa Csetriné Hári

CHINOIN Co. Ltd. Chemical Analytical Development Laboratory

The optimization of a gas-chromatographic test method for intermediates of a drug substance produced by CHINOIN can be seen on the poster.

The aim was to develop a gas-chromatographic method, which can detect and separate all the starting materials, intermediates and their impurities.

It is very difficult to find the optimum conditions for the chromatographic separation of several components. The experimental design was used for this purpose.

The selected method based on the results was validated.

Altogether five GC parameters were considered. The temperature settings and the pressure of the carrier gas were changed the others were kept constant.

2^5-1 = 2⁴ = 16 fractional factor design was applied for the experiments.

The experiments carried out are to be evaluated according to the objective function.

The resolution is a generally used factor for describing the separation between two adjacent peaks in chromatography.

The experiments were evaluated using the Software Statistica 6.0 on the basis of resolutions and the run time.

First the individual objective functions were evaluated one by one.

The equation of the estimated surface can be calculated by taking only the significant main effects into consideration.

The suitability of the fitted model can be controlled by the experiments performed in the central point and by the analysis of the residuals.

The aim of the experiments is to select the optimum GC conditions for all the factors, so it is the best solution to choose a multiple objective function, the so-called desirable function.

The evaluation of the d function shows which parameters have main effect for the chromatographic separation and which resolutions are the most sensitive for the changing of parameters.

After the execution and evaluation of experiments the available information is sufficient to set up the optimum conditions of the chromatographic method, but further experiments are necessary to estimate the parameters of the function for the exact description of the process.