Ab initio study of doped carbon nanotube sensors

Recently great advances have been made in demonstrating the viability of using carbon nanotubes (CNTs) to detect the presence of chemical gases such as NO2, NH3, and O-2, and they have led to the design of a new breed of sensor devices. Based on intrinsic CNTs, the devices are capable of detecting small concentration of molecules with high sensitivity under ambient conditions. However, these devices have a limitation that only molecules binding to a carbon nanotube can be detected. They are currently limited to NH3, NO2, and O-2, and a host of highly toxic gases (such as carbon monoxide), water molecules, and biomolecules cannot be detected using these intrinsic CNT devices. Recent efforts on externally functionalizing CNT surface and internal doping in CNT only result in temporary sensing capability due to the weak van der Waals interaction between CNT and doped materials. In this paper, we propose the concept of a new type of nanoscale sensor devices that can detect the presence of CO and water molecules. To overcome the reliability problem, these devices are developed by substitutional doping of impurity atoms (such as boron, nitrogen atoms) into intrinsic single-wall carbon nanotubes or by using composite BxCyNz nanotubes. Using first-principle calculations, we demonstrate that these sensor devices can not only detect the presence of CO and water molecules, but also the sensitivity of these devices can be controlled by the doping level of impurity atoms in a nanotube.