DESIGN CRITERIA OF A GAS-CHROMATOGRAPHY MASS-SPECTROMETRY BASED EXPERT SYSTEM FOR ARSON ANALYSIS

An analysis system is presented that is specifically geared toward the detection of petroleum based volatiles in suspect arson samples. Advantage is taken of the high selectivity of mass spectrometry toward aliphatic and aromatic hydrocarbons. Data acquired by gas chromatography-mass spectrometry are subjected to automated post run data treatment. The system attempts to categorize chromatographic profiles by establishing the presence of key components. Comparison is made to data from standard stored in the memory of the computer. Retention time information as well as mass spectral criteria are used. Based on the detection of diagnostic components and determination of their relative intensities, samples are categorized as falling into one of five categories of petroleum based accelerants. Extensive use is made of Macro programming. The rationale of the search routine is discussed and details of the program structure are provided. A typical example is presented.