A foundational study of fire debris interpretation using quantitative measures of chromatographic features in gasoline and the use of graphical display to demonstrate data sufficiency

Abstract

The analytical process for identifying ignitable liquids is based on fundamental chemical properties; however, the current interpretation of these properties as chromatographic data relies on subjective pattern recognition techniques. The subjectivity of these pattern recognition techniques increases with the presence of complex matrix contribution. To make the fire debris interpretation process more standardized and objective, a novel method is proposed for analyzing fire debris Gas Chromatography-Mass Spectrometry (GC–MS) data using quantitative measures of chromatographic features of interest. These features are represented by peak height ratios observed in the Total Ion Chromatograph and Extracted Ion Profiles.Chromatographic features of interest in 150 gasoline samples were evaluated and 64 chromatographic peak height ratios were selected for study. Statistical analysis was conducted to determine the variation observed for each of these ratios in the gasoline samples and to determine the frequency of these features in negative matrix samples. This information was evaluated to determine relative significance, as represented by the assigned points for each of these features. When summed and used as plot values, these cumulative scores graphically display the separation of gasoline samples from negative matrix samples using this methodology. The scores were used to create a sufficiency graph, which is a graphical display detailing the totality of data supporting a potential gasoline identification. The sufficiency graph also identifies the “gray” area where analysts are more likely to form differing opinions.The methodologies introduced are a step toward a documentation process that ensures greater transparency in fire debris examinations and comparisons. The methods generate a quantitative sufficiency graph for consistent data interpretation and documentation.