Date of Award


Degree Type



The increased number of terrorist attacks using improvised explosive devices (IEDs) over the past few years has made the trace detection of explosives a priority for the Department of Homeland Security. Considerable advances in early detection of trace explosives employing spectroscopic detection systems and other sensing devices have been made and have demonstrated outstanding performance. However, modern IEDs are not easily detectable by conventional methods and terrorists have adapted to avoid using metallic or nitro groups in the manufacturing of IEDs. Instead, more powerful but smaller compounds, such as TATP are being more frequently used. In addition, conventional detection techniques usually require large capital investment, labor costs and energy input and are incapable of real-time identification, limiting their application. Thus, a low cost detection system which is capable of continuous online monitoring in a passive mode is needed for explosive detection.

In this dissertation, a thermodynamic based thin film gas sensor which can reliably detect various explosive compounds was developed and demonstrated. The principle of the sensors is based on measuring the heat effect associated with the catalytic decomposition of explosive compounds present in the vapor phase. The decomposition mechanism is complicated and not well known, but it can be affected by many parameters including catalyst, reaction temperature and humidity. Explosives that have relatively high vapor pressure and readily sublime at room temperature, like TATP and 2, 6-DNT, are ideal candidate for vapor phase detection using the thermodynamic gas sensor. ZnO, W2O3, V2O5 and SnO2 were employed as catalysts. This sensor exhibited promising sensitivity results for TATP, but poor selectivity among peroxide based compounds.

In order to improve the sensitivity and selectivity of the thermodynamic sensor, a Pd:SnO2 nanocomposite was fabricated and tested as part of this dissertation. A combinatorial chemistry techniques were used for catalyst discovery. Specially, a series of tin oxide catalysts with continuous varying composition of palladium were fabricated to screen for the optimum Pd loading to maximize specificity. Experimental results suggested that sensors with a 12 wt.% palladium loading generated the highest sensitivity while a 8 wt.% palladium loading provided greatest selectivity. XPS and XRD were used to study how palladium doping level affects the oxidation state and crystal structure of the nanocomposite catalyst.

As with any passive detection system, a necessary theme of this dissertation was the mitigation of false positive. Toward this end, an orthogonal detection system comprised of two independent sensing platforms sharing one catalyst was demonstrated using TATP, 2, 6-DNT and ammonium nitrate as target molecules. The orthogonal sensor incorporated a thermodynamic based sensing platform to measure the heat effect associated with the decomposition of explosive molecules, and a conductometric sensing platform that monitors the change in electrical conductivity of the same catalyst when exposed to the explosive substances. Results indicate that the orthogonal sensor generates an effective response to explosives presented at part per billion level. In addition, with two independent sensing platforms, a built-in redundancy of results could be expected to minimize false positive.