M. Doshi, E.P. Fahrenthold
University of Texas at Austin,
Keywords: carbon nanotubes, sensors, explosives
Summary:Gas phase detection of explosive molecules is a challenging task, since low vapor pressures and common functional groups require both sensitive and selective detection systems. Carbon nanotube (CNT) based detection systems have shown sensitivity to particular analytes but often lack selectivity. CNT functionalization has been studied as a means to increase the sensitivity and selectivity of nanocarbon based sensors. Experiments normally employ mixes of metallic and semiconducting CNTs, due to the high cost of segregating metallic and semiconducting CNTs. In this research, the chemiresistive sensing performance of both metallic and semiconducting CNTs has been evaluated using ab initio modeling. The metallic and semiconducting CNTs were functionalized using the four most commonly occurring functional groups: hydroxyl, carboxylic acid, oxygen, and amine groups. The chemiresistive sensor responses have been modeled for one nitroaromatic explosive (TNT) and two nitramines (HMX and RDX). Analysis of CNT sensor arrays consisting of functionalized metallic or functionalized semiconducting CNTs indicates that they show different sensitivities to the modeled explosives, and are also complementary. Principal component analyses of the modeled sensor arrays indicates that both CNT types are individually capable of distinguishing: (1) common background gases from both nitroaromatic and nitramine explosives, (2) nitroaromatic from nitramine explosives, and (3) similar nitramine explosives from each other. The modeling results indicate that segregated CNTs show excellent potential for the development of large sensor arrays which are also compact and have low power requirements.