Transfer of retention patterns in gas chromatography by means of response surface methodology

Accurate transfer of retention patterns in temperature-programmed gas chromatography is challenging because minor variations in column properties and experimental conditions may have significant impact on the elution patterns. An experimental method for accurate transfer of retention indices is proposed and validated. The methodology is based on response surface methodology and experimental design. The temperature rate and the start temperature of the rate are varied systematically in the region where the optimal conditions are expected to be found. Response surfaces that explain the absolute deviation to the target retention indices are calculated for each compound. These response surfaces are thereafter averaged and the minimum in the average surface is regarded as optimal conditions for reproduction of the retention pattern. The methodology was applied on fatty acid methyl esters using equivalent chain lengths as the retention index system. Two different target patterns were tested on two BPX-70 columns with different dimensions. Validation of the proposed conditions showed that the retention patterns could be reproduced with an error that was only fractions of a peak width.