A novel protocol for producing low-abundance targets to characterize the sensitivity limits of ignitable liquid detection canines

Abstract Ignitable liquid detection canines have an extraordinarily sensitive sense of smell. Previous studies estimating canine sensitivity to ignitable liquids have been unable to demonstrate a level below which canines will not indicate ignitable presence. Forensic laboratories provide corroborating evidence and explain reasons for canine indications. Laboratories provide protection against false-positives and greater context for the exhibit contents to both investigators and the courts. However, the methods routinely used in forensic laboratories lack the sensitivity required to match canine sensitivity estimates, leading to frustration when canines provide a strong indication, but the laboratory findings are negative. In order to develop a laboratory method that can match the sensitivity of canines to trace ignitable liquids, an estimation of canine sensitivity is required. We present a protocol for producing extremely clean, porous tile substrates, and a scheme for quantitatively producing target samples with any desired abundance by using a solvent that does not elicit a response from the canines. Search experiments with two canine-handler teams validate the methods and provide the first estimate of the ignitable liquid detection canine sensitivity limit for gasoline: both canines repeatedly located gasoline at 5 pL but provided no indication at any level ≤3 pL. The protocol allows standardization of substrates and targets for canine olfactory testing, while the estimated canine sensitivity to gasoline provides a preliminary target for forensic laboratories performing confirmatory analysis on exhibits selected by ignitable liquid detection canines.