Flawed explosives detection: Sniffer dogs fail to detect C-4 components
- Published: Jan 9, 2014
- Author: Steve Down
- Channels: Gas Chromatography
Training sniffer dogs for explosives
Image: British Army
The use of dogs to sniff out explosives is becoming more common and relies on extensive training programs to ensure that the animals can detect low levels of the explosives. In some cases it is clear that it is not the explosive compound itself that is being recognised, but the accompanying components of the explosive mixture. This is the case for RDX, a nitro-based explosive which has a very low vapour pressure and is difficult to detect at room temperature.
So, for the explosive C-4, which is based on RDX, the headspace is dominated by volatile components like solvents, binders, impurities and additives. In practice, very little is known about which compounds, or combinations of them, the sniffer dogs respond to. Some people subscribe to the bouquet theory in which dogs recognise a characteristic mixture of components from the explosive formulation. However, recent research has suggested that the dogs respond most strongly to the volatile compounds on which they were trained.
Clearly, factors like this must be taken into account when devising the canine training programs so that the dogs act as efficiently as possible in the field. There might also be legal implications if the scientific basis of the canine detection program is not fully established.
For these reasons, John Goodpaster and colleagues from Indiana University-Purdue University Indianapolis (IUPUI) decided to take a closer look at the canine detection of C-4. They based their initial studies on three components of the C-4 formulation which have been proposed as training aids. 2,3-Dimethyl-2,3-dinitrobutane (DMNB) is present as a compulsory component under US law. The second compound was cyclohexanone, a residual solvent from the manufacture of RDX that is present in many non-explosive materials and could lead to false alarms.
The team also examined 2-ethyl-1-hexanol (2E1H) which is a major volatile component of C-4 and is found abundantly in the headspace. There have been some reports that 2E1H is emitted from other materials, which could lead to false positives during canine detection, so their first step was to examine a wide range of plastic articles to see if this was the case.
Interfering sources of 2-ethyl-1-hexanol
In the first instance, the team confirmed that 2E1H was a hydrolysis product of the plasticiser bis(2-ethylhexyl) adipate which is a component in the C-4 formulation. Then they tested 26 everyday items as potential 2E1H sources. They included PVC tiles, PVC pipes, electrical tape, plastic food wrapping, bubble wrap, lottery and movie tickets, a playing card and a shower curtain.
Portions of each were sampled by SPME using a polydimethylsiloxane fibre which was the best performer from three types of coatings. The fibres were then desorbed in the inlet of a GC/MS instrument and the volatile compounds were analysed by electron ionisation.
A total of 11 items emitted what the researchers regarded as “appreciable levels” of 2E1H: five types of PVC tile, two types of PVC pipe, three brands of electrical tape and a credit card. Bearing in mind that many more plastic items are in wide circulation and may also emit 2E1H, they recommended that the use of 2E1H as a training aid should be reconsidered.
Canine field trials reveal inadequacies
The team then organised their own canine field trials using 33 certified dogs and their handlers that had been trained to detect C-4 by their own law enforcement agency. They were used to detect C-4 and the three odour compounds selected above plus the plasticiser bis(2-ethylhexyl) adipate. Samples of C-4 or of each compound were added to special sniffer tins which were placed in perforated paint cans.
The detection rate of C-4 was lower than expected at 67% and the team suggested that this might be due to the unfamiliarity of the dogs for the type of containers used. However, when canine teams are deployed in the field, the suspect samples could be in any type of container and not conveniently placed in tins, so this should not be a factor.
More surprisingly, only one pass by one dog detected 2E1H. The rates for DMNB and cyclohexanone were better at 30 and 33%, respectively, but are still relatively low. There was also a lack of correlation between the dogs’ ability to detect C-4 and bis(2-ethylhexyl) adipate.
The poor responses suggest that dogs that are trained to detect C-4 using these particular individual compounds will perform poorly and deliver a poor detection rate. It is possible that they might react better to a mixture of 2E1H, cyclohexanone and DMNB, although this theory was not tested. Alternatively, they should be trained on C-4 itself. Either way, the team suggest that the single compound training system for explosives sniffer dogs should be re-evaluated.
Forensic Science International 2014 (Article in Press): "On the smell of composition C-4"
Article by Steve Down
The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.