Agneta Bemgård

Gas chromatographic analysis of high-molecular-mass polycyclic aromatic hydrocarbons: II. Polycyclic aromatic hydrocarbons with relative molecular masses exceeding 328

Abstract Three columns were used for the gas chromatographic analysis of polycyclic aromatic hydrocarbons (PAHs) with relative molecular masses (Mr) up to 450. Two of the columns were commercially available, coated with a 50% methyltrifluoropropyl-substituted polysiloxane a 5% diphenyl-substituted methylpolysiloxane. The third column was laboratory made, coated with a biphenyl-substituted silarylene-siloxane copolymer. All three columns were utilized for the analysis of high-Mr PAHs as regards both thermal stability of the stationary phases, i.e., low bleeding rate, and chromatographic efficiency. The column coated with a trifluoropropyl-substituted stationary phase showed, however, a low separation efficiency, possibly owing to low solute stationary phase compatibility. The biphenyl-substituted stationary phase, on the other hand, showed a very high separation efficiency, but the retention of the PAHs was significantly higher on this column compared with the other two, leading to the demand for higher oven temperatures. Different retention mechanisms were observed on these columns, as shown by differences in the retention indices of the PAHs measured in a system using PAHs as retention index markers. A comparatively faster elution of non-planar PAHs was observed on the columns coated with the trifluoropropyl-substituted stationary phase and the biphenyl-substituted stationary phase compared with the column coated with the 5% diphenyl-substituted polymer. The usefulness of the columns for separations of high-Mr PAHs is demonstrated by gas chromatograms of carbon black extracts and a coal tar extract standard reference material.

Gas chromatographic analysis of high-molecular-weight polynuclear aromatic hydrocarbons: I. Molecular weight 328

Abstract Various columns were investigated with respect to the gas chromatographic analysis of polynuclear aromatic hydrocarbons (PAHs) having molecular weights that exceed 300. Among the columns tested, four columns differing in stationary phase composition were carefully evaluated. Three of these columns were commercially available. The evaluation was focused on cata-condensed PAHs of molecular weight 328. Temperature-programmed retention indices with PAHs as reference markers were calculated and compared.

On the relation between retention indexes and the interaction between the solute and the column in gas-liquid chromatography

Gas−liquid chromatography retention indexes for organic molecules are determined by the interaction between the molecule and the column liquid phase. In this article, a model for calculating the interaction energy between a molecule and a dielectric wall is developed. The model is at least to our knowledge the first attempt to predict retention indexes from the interaction between the molecules and the column. This approach to predict retention indexes is radically different from methods proposed before. Earlier predictions of the retention indexes have been done by a large number of descriptors, which were linearly correlated to the retention indexes. The developed model has been tested for polycyclic aromatic hydrocarbons mainly with a molecular weight of 302. For the molecules with MW 302 the obtained correlation coefficient is 0.92. A somewhat simpler model is used to fit PAH with different MWs. A correlation coefficient of 0.998 is obtained if the retention indexes were fitted to the logarithm of the...

Universal set-up for measurement of diffusion coefficients in supercritical carbon dioxide with flame ionization detection

Abstract Improvements to the chromatographic apparatus for diffusivity measurements in supercritical carbon dioxide with flame ionization detection have been designed and evaluated. Modifications concerned the injection with partial solvent evaporation in the injector loop, separation of solutes and solvent in a short precolumn coated with a thin film of PS 264 stationary phase, and a back-pressure device. Diffusion coefficients of C15-C18n-alkanes, dodecanone, pentadecanone, methyl myristate and biphenyl in supercritical carbon dioxide at 100°C and 125°C at 100 atm were determined. The method of optimization of Dg to a plot of H vs. v was evaluated for the estimation of diffusion coefficients. The estimated diffusion coefficients have been compared with the average diffusion coefficients and diffusion coefficients calculated from the Wilke-Chang equation. Relatively good agreement of measured diffusion coefficients and those calculated from the modified Wilke-Chang equation has been obtained for biphenyl and polar solutes.

Assessing breakthrough times for denuder samplers with emphasis on volatile organic compounds

Abstract This paper presents a theoretical background useful for the prediction of cylindrical denuder sampler performance, applied to volatile organic compounds (VOC). Two methods are presented that can be used together with any type of adsorption isotherm. One computer method is based on a simple recursive modelling of the Gormley-Kennedy equation with a time-dependent denuder length. The other model uses an equation derived directly from the Gormley-Kennedy expression. In practice, the latter, simpler model can readily be used. The most crucial part in the breakthrough assessment is associated with modelling the adsorption isotherm. Two types of models have been considered: the Langmuir and the Freundlich adsorption isotherms. In order to facilitate the comparison between samplers with different adsorption capacities and different physical dimensions, a reduced length versus time plot is introduced.

Gas chromatographic methods for the assessment of binary diffusion coefficients for compounds in the gas phase

Abstract Binary diffusion coefficients, Dg, were calculated for a number of compounds by using three gas chromatographic methods. All three methods showed good agreement when Dg was extrapolated to zero gas velocity. For the compounds tested, all Dg values were lower than those achieved by using the Fuller—Schettler—Giddings equation.