Carbon nanotubes for sensing and terahertz applications

Abstract

Emerging Internet of Things (IoT) requires reliable and high-performance sensors. Carbon nanotubes (CNTs) have been proposed as a promising candidate for the next-generation sensors due to their unique properties. CNTs have high surface to volume ratios and, therefore, they are especially suited for interaction with external stimuli, such as molecules and electromagnetic waves, to transduce electronic signals. The properties of CNTs depend on structural parameters and on the growth technique and affected by encapsulation and substrates, as well as dopants. Besides the individual forms of CNTs, the network system of randomly oriented CNTs exhibit considerable potential for sensing applications. In particular, percolative behavior, which is a dramatic change in the electrical conductance of the CNT cluster near the percolation threshold, where long range electrical path is established allows for an additional adjustment and optimization of the sensing properties. Despite of a number of experimental and theoretical demonstration of CNT sensors, a fundamental issue for further performance improvement is yet to be solved: inherent randomness in sizes, types, alignment, and surface properties of the carbon nanotubes. In this thesis, we focus on the randomness control strategies for CNT field-effect transistors (FETs) and CNT network-based sensors. We report on the compact unified charge control model (UCCM) for single CNT based FETs to enable accurate simulation of plasmonic THz response and temperature dependent electrical conductance in the CNT FETs, and thereby, to allow optimization of the sensing performance of CNT FETs. We also propose a method to enhance the sensing performance of CNT random networks near the percolation threshold. We demonstrate the substantial conductance change of the CNT network in response to temperature variance and THz radiation near the critical CNT density. These results can provide a novel path to develop feasible methods for improving the CNT sensor performance.