Characterizing the role of defects on the sensing performance of carbon nanotube and graphene based gas sensors

carbon nanotubes;graphene;chemical sensor;defects;electron hopping;network analysis

Lin, Kevin Y.

关键词: carbon nanotubes;    graphene;    chemical sensor;    defects;    electron hopping;    network analysis;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/26367/Lin_Kevin.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

The major theme of this thesis is to investigate the influence of defects on the sensing mechanisms and performance of gas sensors made from films of carbon nanotubes, chemical vapor deposition (CVD) graphene film, CVD graphene ribbons, and surfactant exfoliated graphene. The definition of defects on carbon materials is the disordered carbon atoms formed due to dislocations, vacancies, and deformations. These defects were introduced during processing. The thesis is separated into three sections that analyze various types of defects on these carbon based chemical sensors. First section focuses on single-walled carbon nanotubes with point defects. It will be demonstrated that these point defects can alter the main sensing mechanism of the CNT based sensors. There is a controversy in literature on whether the sensing response seen in carbon nanotube chemiresistors is associated with a change in the resistance of the individual carbon nanotubes or changes in the resistance of the junctions. A network analysis was carried out to better understand the relative contributions of the carbon nanotubes and the junctions on the change in resistance of the carbon nanotube network. It was found that the dominant mode of detection in carbon nanotube networks changes according to the defect level in the carbon nanotubes which may explain the apparently contradictory results in the literature. With high concentration of the point defects in the carbon nanotube film along with applying high electric field, the Poole-Frenkle conduction regime can be induced. Generally, desorption of gases from carbon nanotubes is a slow process that limits the carbon nanotubes’ utility as sensors. It will be demonstrated that electron flows in the carbon nanotube above the Poole-Frenkel conduction threshold can stimulate adsorbates to desorb without heating the sensor significantly. This desorption process is analogous to electron stimulated desorption, but with an internally conducted rather than externally applied source of electrons. As a result, this gives a fast and reversible CNT sensor within seconds. An application can be utilized to use carbon nanotubes based GC detector for multi-component chemical analysis. The approach is to use a CNT based detector in a series configuration with a gas chromatography column. It will demonstrated that a mixture of nine different compounds can be detected with these CNT based detectors when the detector operates in current stimulated desorption (CSD) mode. This is the first demonstration of a CNT based GC detector to analyze multi-component gas mixtures providing a new sensing approach for online air quality control and health monitoring applications. The second section of the thesis focuses on analyzing two-dimensional line defects arises from wrinkles and grain boundaries as well as edge defected created manually on the graphene film synthesized with chemical vapor deposition (CVD) process. Due to two-dimensional nature of the graphene film, adsorption of isolated analyte molecules on point defects has minimal effect on graphene resistance because current pathways can always form around the adsorbate. In contrast, analytes adsorbing on line defects lead to significant changes in resistance. It will be demonstrated that polycrystalline graphene easily obtained through chemical vapor deposition contains line defects. This can offer a scalable path to 50x more sensitive chemiresistors than mechanically exfoliated crystalline graphene. Moreover, current flow can be confined by cutting the polycrystalline graphene into ribbons, the sensitivity increases by another factor of four. These results show that polycrystalline graphene has extraordinary sensitivity, achieved through geometry and linear defect density. The last section of the thesis focuses on intrinsic two-dimensional edge defects on graphene isolated with sodium cholate surfactant assisted exfoliation of graphite powders. With this technique, it is possible to produce high edge defect concentrations for individual graphene island due to its micron-sized dimension offering more defects per unit area. Various randomly-stacked oxide-free graphene films can be formed with various filtration volumes. It will be demonstrated the films produced can range 6 orders of magnitude in film conductance. At thinner graphene films, the electron transport mechanism is mainly through two-dimensional variable range hopping. At thicker graphene films, electron transport is through fluctuation-assisted tunneling. It is also observed the graphene films go from semiconducting-like to metallic-like behavior at around 8 mL filtration volume. At low filtration volumes, oxide-free randomly stacked graphene film sensors showed better sensitivity towards target molecules compared to polycrystalline/ribbon graphene and defective CNT gas sensors due to two-dimensional electron hopping. As the graphene film thickness increases, there is a shift in conduction mechanism from two-dimensional electron hopping to metallic-like conduction which explains the drop in sensitivity as the filtration volume increases.