Multidimensional gas chromatography of oxidative degradation products in algae-derived fuel oil samples using narrow heartcuts and rapid cycle times

Abstract

To characterize a fuel's thermal and storage stability an understanding of the process of oxidation and oxidation pathways is essential. Oxidation pathways commence with hydroperoxides which quickly decompose to form a range of alcohols, acids and other oxygen-containing species. In the presence of significant levels of hydrocarbon-based matrix, analysis of these heteroatomic species is difficult. Applying multidimensional gas chromatography with very narrow heart-cut windows (0.20 min) minimizes the number of compounds transferred to the second dimension ( 2D) column during each heart-cut. Successive heart-cuts every 2.00 min are taken throughout the analytical run, since each heart-cut has a maximum retention on 2D of <2.00 min on the fast elution 2D column. Subsequent analyses involve incrementing or offsetting the heart-cut windows by 0.20 min, so after 10 analyses, a complete coverage of the sample components can be obtained. On the polar 1D and non-polar 2D phase column arrangement, non-polar matrix compounds elute last on the 2D column, and this determines the largest 2 t R; i.e. 2 t R < P M to ensure retained components on 2D will not overlap with subsequent heart-cuts. Heartcutting is supported by cryotrapping at the start of the 2D column in order to provide significantly better resolution. Good quality MS library match data generally demonstrate the high resolution separation of oxygenates achieved. Whilst 1D GC–MS was unsuccessful in identifying any of the oxygen-containing compounds reported here, good correlation of MS data (with average MS library similarity data) for acids (903), alcohols (909), ketones (941) and aldehydes (938) in the sample is obtained. The method requires ten sequential runs, and this can be accomplished automatically once the events table is set up. However if fewer target compounds are to be transferred, a reduced number of sequential runs can be implemented.