Chemical Characterization Of Aviation Fuels By Gas Chromatography-Mass Spectrometry
Abstract
This research proposes that analysis of greater depth can be performed on fuels to help understand their composition better than what is currently performed. At present, there is a disconnect between the groups that use fuels and the groups that make fuels-neither side knows precisely what the other wants. By having a better understanding of a fuel's chemical makeup, correlations between different components and different properties may be found. Using this knowledge, fuels could be made exactly how an end-user would need them. This could reduce waste of undesirable fuel as well as drive us away from fossil fuel dependence since synthetic generation of these fuels will likely be more beneficial when we fully understand how different components of fuel affect its different properties. Using these ideas, I have developed a method to characterize the different hydrocarbon types present in an aviation fuel as well as attempt to determine the degree of branching in its paraffinic components. As paraffins typically make up the majority of most fuels, this takes a great step toward better understanding the makeup of an aviation fuel. This method, while similar to ASTM methods that characterize the hydrocarbon types present in different cuts of fuel, refines those methodologies for aviation fuel specifically and expands upon them with branching determinations. Understanding the branching present in hydrocarbons will provide a better understanding of the combustion properties and other physical properties including things like volatility. All of this together means that this facet of information will better allow for prediction of fuel properties from knowledge of fuel composition. This would save time manually testing things like freezing point, heats of combustion, and so forth. In this paper I walk through the way in which this method was formed and the method is applied in the characterization of a JP-8 aviation fuel sample. While this analysis will show that the proposed branching determination does not work, I believe that the trends do exist in the mass spectral data by which to perform this determination.