Capillary Chromatography - The Analysis of Gasoline 1



However, it does appear more suitable for the resolution of larger molecules. Again, the composition of the test mixture provides information on the general retentive character of the stationary phase. The elution of n-pentadecane in only about 17 minutes at a temperature of only 125ûC would indicates that the dispersive contributions by the stationary phase to solute retention were relatively minor. In contrast, the C8 ethylhexanoic acids are eluted just before the n-pentadecane exposing the very strong polar character of the phase. The isomeric dichlorobenzenes demonstrate a good selectivity for the spatial isomers and there is also an effective separation of the two enantiomers. The separation of the dichloro-benzenes again indicates that the presence of strong induced polar interactions between the aromatic nuclei of the solutes and the strong polar groups of the stationary phase. Similarly, the strong polar interactions that are seen to occur between the carboxyl group of the enantiomer that penetrates closest to the stationary phase, and the neighboring polar groups of the interactive site obviously provide the necessary selectivity to resolve it from the other, slightly more excluded, enantiomer.


The Analysis of Gasoline


Gasoline is a multi-component hydrocarbon mixture of which many of the components have very similar molecular weights and all have interactive properties that are almost exclusively dispersive (with the exception of the aromatic hydrocarbons that can interact by induced dipoles if the stationary phase is strongly polar). The structure of many of the hydrocarbons components are also very similar and the mixture contains many isomers. It follows that the separation factors between individual components are likely to be very small and, consequently, to achieve a separation, columns of very high efficiency will be essential.


Open tubular columns are ideal for this type of separation and, in fact, it is impossible to separate gasoline efficiently with a packed column, (and that will be true even if the column is 50 ft long and the inherent long analysis times could be tolerated). As the separation demands a large number of theoretical plates a relatively small diameter open tubular columns must be used. Furthermore, several hundred thousand theoretical plates will be necessary and so the column must also be fairly long.