Predicting Gas Chromatographic Retention Times using Thermodynamics.

摘要

An investigation into different methods to determine thermodynamic parameters from gas chromatographic retention data for their application in predictive models was performed. Based on a rigorous evaluation it was determined that a linearized three-parameter model was most suitable for this application. These thermodynamic parameters were found to vary when the carrier gas was changed from helium to hydrogen or nitrogen. This effect could not be attributed to the kinetics of the system.;Quantitative structure retention relationships were developed to predict retention times for a series of alkanes based on thermodynamic data from approximately 75 compounds. Although not as accurate of predictions were obtained compared to the additive model these methods do show promise.;An additive model was constructed which was able to predict the retention times of a series of alcohols and ketones in both isothermal and temperature-programmed modes with an average error of less than 10 seconds in isothermal mode and 5 seconds in temperature-programmed mode.