Abstract
The analysis of residual accelerants in fire debris is commonly carried out by a three‐step procedure: sample preparation; separation and detection; and data interpretation. Each of these steps can be optimized individually but successful analysis requires that they are compatible with each other. The isolation of residual accelerant from fire debris requires that several methods are used to cover the range of fueis that are commonly used by arsonists. Since almost all incendiary fires are set with petroleum based fuels such as gasoline or heating oil, analysis is targeted toward hydrocarbons. Capillary column gas chromatography on apolar phases is now the overwhelmingly predominant method of separation. Data interpretation is commonly carried out by visual comparison of chromatograms.
Fire debris analysis presents some unique challenges that are not often encountered in other fields. The analyte may be present at only trace levels and pyrolysis products from building materials or furnishings may dominate chromatographic patterns. Synthetic polymers may act as precursors to hydrocarbons that compete with substances typically found in petroleum based fuels. Exposure to heat and other environmental factors may also generate severe distortions in the chromatographic profiles of accelerants.
Unfortunately, there is no simple solution to these problems. Methods are available to reduce some of the chemical noise introduced by interferences and thus enhance the recognizability of the target substances. Mass spectrometry, in combination with a modern data system, is the most effective approach to filter out unwanted substances. The interpretation of the analytical results is aided by scaling, side‐by‐side comparison, or stacking of chromatograms. An additional advantage of such computerized systems is the possibility of complete automation of the analysis.
In this communication, we look at the interplay of chromatographic resolution, noise reduction by mass spectrometry, and automated data evaluation. Examples from model experiments and from simulated arson samples are presented.