A chromatographic method has been developed for the analysis of petroleum liquids to replace the standard distillation which is used by the analytical laboratories to split the sample into light and heavy fractions, thus permitting the detailed characterization of the former.
The microdistillation, which uses only a few microlitres of sample, is based on a dynamic accelerated vaporization of the light components assisted by an overlying flow of an inert gas and it achieves a fast and efficient separation between the residue and the distilled product. As the microdistillation unit is physically attached to the gas analyser, it thereby constitutes a closed system with the latter and allows the vaporized fraction to be characterized directly by gas chromatography. The remaining residue is subsequently characterized by liquid chromatography.
The accuracy of the method was tested against synthetic mixtures of known composition and against stock tank oils. The microdistillation can be extended to the analysis of all types of multicomponent mixtures with constituents exhibiting a wide boiling point distribution.
INTRODUCTlON
Gas chromatography (GC) is now the universally accepted method for charactcrizing light hydrocarbon mixtures. The use of long wall-coated capillary columns with the equivalent of hundreds of thousands of theoretical plates permits the quantitation not only of single carbon number groups, but also of isomers and hydrocarbons belonging to different homologous series.
GC has been used for a long time to simulate the distillation of petroleum products and to provide plots of amount distilled WV-sus boiling point. Eggerston et al. ' in 1960 and Green et ~1.~ in 1964 described methods which they called "simulated distillation", using low-resolution GC with non-polar columns, to evaluate the boiling point distribution of hydrocarbon samples. The simulated distillation method is based on the assumption that individual components of a sample are eluted from a chromatographic column in the order of their boiling points. Because of the regularity