Effects of Membrane Properties on the Odor Emanating from Training Aids for Explosive-Detective Canines: An Interview with John Goodpaster and Himanshi Upadhyaya

Many previous studies have shown that canines rely on the odor of volatile organic compounds that are emitted by explosive formulations. It is then especially important that canines be trained on authentic odors that are neither degraded nor contaminated. Research by Himanshi Upadhyaya and John V. Goodpaster of the Department of Chemistry and Chemical Biology, Indiana University Indianapolis has focused on using an odor-permeable membrane device (OPMD) to determine any changes in the odor of dynamite and have utilized gas chromatography–mass spectrometry in negative chemicalionization mode in all analysis. Upadhyaya and Goodpaster responded to questions from LCGC International regarding their research.

Briefly discuss the historical benefits of utilizing canines in explosive detection and their most favorable characteristics.
Historically, canines have been used as hunting dogs due to their ability to detect scent for over 12,000 years. The first evidence of the use of canines by police dates to 1888, when bloodhounds were used for tracking the infamous criminal Jack the Ripper in England. In the United States, the first instance of canine training for detecting explosives was in an academic setting at the University of Mississippi in 1971. This study led to the development of the National Explosives Detection Canine Team Program by the Federal Aviation Administration (FAA) in 1972. Finally, in 1986, the Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF) initiated the training of accelerant detection canines (ADCs), which further led to the training of ATF explosive detection canines (EDCs).

EDCs are obtained from canine populations that have undergone centuries of selective breeding for purposes such as hunting, herding, and protection. Canines can be regarded as a highly mobile sampling system coupled with extremely sensitive odor detection capabilities. Furthermore, canines can be trained to be extremely selective in detecting odors originating from different explosive types which makes them an indispensable part of safety and law enforcement programs.

You state in your paper (1) that, while recent times have seen an advancement in development and optimization of a plethora of analytical techniques such gas chromatography–mass spectrometry (GC–MS), Raman spectroscopy, infrared (IR) spectroscopy, and electrochemical sensors for explosive detection, you believe that canines offer advantages to these methods. Can you talk about those benefits, and why are the methods still lacking?
The ability of trained canines to focus on a single odor and tune out other odors that contribute to the background “noise” makes them exceptionally selective for use as EDCs. Analytical techniques such as GC–MS, IR spectroscopy, and Raman spectroscopy are typically bulky, slow, and lab-based. This equipment also requires technical expertise to be handled. The “background noise” (the presence of other chemical components in the explosive) can often lead to a complex spectra which would require the need for deconvolution. Electrochemical sensors have a shorter shelf life, and even though they can be sensitive, they can show poor reproducibility. In comparison, canines show high sensitivity and selectivity, can be trained to detect a wide variety of odors, and are mobile. Importantly, canines do NOT have specificity; their detection response is the same, regardless of the explosive. Hence, a canine alert can only be regarded as a presumptive identification.

The study described in this paper focuses on the use of an odor-permeable membrane device (OPMD) to determine any changes in the odor of dynamite. Why was GC–MS your technique of choice in your analysis?
GC–MS is a gold standard for detection of volatile organic compounds in the field of analytical chemistry. GC–MS has an established history of reliable detection of volatile organic compounds (VOCs) present in different substrates including explosives. Negative chemical ionization was chosen as the mode of ionization in this study to produce high sensitivity and selectivity for nitrated explosives.

Briefly state your findings in this study.
A training aid delivery device (TADD) is comprised of a gas-tight chemical polypropylene lid, a glass jar, a polypropylene membrane holder, and an oleophobic and hydrophobic membrane permeable to VOCs originating from the explosive (in this case, dynamite) and impermeable to contaminants from the surroundings. Overall, nine synthetic membranes and six glass fiber membranes were tested to understand the odor availability of dynamite in a TADD. It was found that out of the two known chemical constituents present in dynamite, ethylene glycol dinitrate (EGDN) and nitroglycerin (NG), the relative amount of NG was selectively altered by some synthetic and glass fiber membranes.

Do your findings correlate with what you had hypothesized?
It was surprising for us to see that NG was being selectively removed by these membranes, because all tested membranes were oleophobic and hydrophobic in nature, which made them impermeable and less susceptible to external contaminants. The difference in the volatility of the two compounds and the pore size of the membranes played a significant role in selective alteration of the relative odor of NG in TADDs. We hypothesize that the smaller pore size in some of these membranes coupled with lower vapor pressure of NG in comparison to EGDN made it less permeable and hence resulted in the selective alteration of NG.

Was there anything particularly unexpected that stands out from your perspective?
It was interesting to see a significant difference in the relative odor availability of the two compounds in dynamite, EGDN and NG. The fact that both compounds are VOCs, and yet there is a significant difference in their relative odor, opens room for further studies to be conducted using other commonly encountered explosives, such as composition C4, to understand their odor availability for proper canine training.

Were there any limitations or challenges you encountered in your work?
Once the difference in the relative odor availability of NG was established after repeated experiments for the default synthetic membrane H, the next step was to procure more synthetic membranes that could be used as alternatives. These membranes differed from one another in terms of pore size, membrane material, and membrane thickness. Once all nine synthetic membranes were tested, we went on to test six glass fiber membranes and the membranes that did not show significant differences in the relative odor of NG.

What best practices can you recommend in this type of analysis for both instrument parameters and data analysis?
The use of negative ion chemical ionization (NICI) was crucial is having the sensitivity required. We also took pains to avoid contamination through changing gloves, decontaminating surfaces, using disposable charcoal strips to adsorb analytes, among other steps.

Can you please summarize the feedback that you have received from others regarding this work?
We do know that the manufacturer of the TADD has changed the filter membrane they are using as the default membrane exhibited discrimination. Given how recently the paper has been published, we have not received any additional feedback from the scientific community.

What are the next steps in this research and are you planning to be involved in improving this technology?
We are currently working on sampling methods for explosive residues, but the next steps for the TADD would be evaluating if other explosives with complex odors (such as composition C-4 and smokeless powder, for example) suffer from any discrimination effects for compounds we have not yet evaluated.

Reference

1. Upadhyaya, H.; Goodpaster, J. V. Effect of Membrane Properties on the Odor Emanating from Training Aids for Explosive‑Detecting Canines. Anal. Bioanal. Chem. 2024, 416, 4219–4225. DOI: 10.1007/s00216-024-05359-w

Dr. John Goodpaster received his B.A. in Chemistry from Gustavus Adolphus College in St. Peter, MN.He then attended graduate school at Michigan State University in East Lansing, MI where he earned a M.S. in Criminal Justice and a Ph.D. in Analytical Chemistry.Following post-doctoral studies at the National Institute of Standards and Technology (NIST) in Gaithersburg, MD, he served as a Forensic Chemist with the Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF) Laboratory in Ammendale, MD.Dr. Goodpaster is a Professor in the Department of Chemistry and Chemical Biology at Indiana University – Purdue University Indianapolis (IUPUI). Dr., Goodpaster is also a part of the Forensic and Investigative Sciences (FIS) Program at IUPUI.He teaches in the areas of alcohol and drug analysis as well as trace evidence.Ongoing research projects in Dr. Goodpaster’ s laboratory include the chemical analysis of human hair, coupling gas chromatography and vacuum UV spectroscopy (GC/VUV), using chemometric techniques to classify and associate trace evidence, and explosive-detecting canines.

Himanshi Upadhyaya is a PhD candidate in the Department of Chemistry and Chemical Biology, IU Indianapolis specializing in analytical chemistry. Her research focuses on developing methods to detect and analyze explosive mixtures using instruments such as GC-MS and GC-VUV, under the guidance of Dr. Goodpaster. She earned her M.S in forensic chemistry from University of Derby, UK in 2019. When she is not in the lab, Himanshi enjoys hiking, gardening and volunteering at community science events. She can be reached on LinkedIn at linkedin.com/in/himanshi-upadhyaya-916217228