Preliminary evaluation of a next-generation portable gas chromatograph mass spectrometer (GC-MS) for the on-site analysis of ignitable liquid residues

Abstract

The detection of ignitable liquid residues (ILRs) at a fire scene can be valuable evidence to indicate if the fire may have been deliberately set. In this study, a commercially-available portable GC-MS, the TRIDION-9, was evaluated for the purpose of on-site fire debris analysis. The instrument, which incorporates a low thermal mass capillary GC and a miniaturised toroidal ion trap mass spectrometer, is designed to use solid-phase microextraction (SPME) as the default sampling method. Four ignitable liquids (petrol, mineral turpentine, kerosene and diesel) and seven substrates (nylon, polypropylene and wool carpet, rubber and foam underlay, untreated pine and polyurethane) were considered. After method optimisation, substrates were burned and spiked with 0.1 μL of ignitable liquid (mineral turpentine, kerosene, diesel fuel, and both neat and weathered unleaded petrol). Sampling was performed via passive headspace SPME adsorption over 3 min immediately followed by analysis using the TRIDION-9 (analysis time < 2 min). Field trials were performed on fire debris samples collected from an accelerated structural fire. The data from the portable GC-MS analyses permitted the detection and correct classification of ILRs in the majority of samples analysed. Overall, the TRIDION-9 was found to be well-suited to the field-based analysis of fire debris samples as a preliminary screen to provide rapid presumptive results for fire investigators.

Keywords:

• fire investigation
• arson
• accelerants
• portable instrumentation
• hydrocarbons
• presumptive screening

Acknowledgements

The authors would like to thank Phillip Etienne and Joel Walton, Fire and Rescue New South Wales, for their involvement in the project including the provision of fire debris samples for analysis. Canine handler Timothy Garrett, Fire and Rescue New South Wales, is acknowledged for his handling of the accelerant detection canine, Sheba, during the research exercise in Gilgandra. Thanks also to Laura Bowen and Phillipa Nolan, Forensics, Australian Federal Police, for their support and general advice. The staff at Torion Technologies Incorporated (Utah, USA) provided valuable assistance including instrument troubleshooting when required. Finally, the authors would like to thank David Royds, National Centre for Forensic Studies, University of Canberra, for his assistance with the data analyses.

Disclosure statement

No potential conflict of interest was reported by the authors.