Effects of low energy E-beam irradiation on graphene and graphene field effect transistors and raman metrology of graphene on split gate test structures

Apart from its compelling performance in conventional nanoelectronic device geometries, graphene is an appropriate candidate to study certain interesting phenomenon (e.g. the Veselago lens effect) predicted on the basis of its linear electron dispersion relation. A key requirement for the observation of such phenomenon in graphene and for its use in conventional field-effect transistor (FET) devices is the need to minimize defects such as consisting of -- or resulting from -- adsorbates and lattice non-uniformities, and reduce deleterious substrate effects. Consequently the investigation of the origin and interaction of defects in the graphene lattice is essential to improve and tailor graphene-based device performance. In this thesis, optical spectroscopic studies on the influence of low-energy electron irradiation on adsorbate-induced defectivity and doping for substrate supported and suspended graphene were carried out along with spectroscopic and transport measurements on graphene FETs. A comparative investigation of the effects of single-step versus multi-step, low-energy electron irradiation (500 eV) on suspended, substrate supported graphene and on graphene FETs is reported. E-beam irradiation (single-step and multi-step) of substrate-supported graphene resulted in an increase in the Raman ID/IG ratio largely from hydrogenation due to radiolysis of the interfacial water layer between the graphene and the SiO2 substrate and from irradiated surface adsorbates. GFETs subjected to single and multi-step irradiation showed n-doping from CNP (charge neutrality point) shift of ∼ -8 and ∼ -16 V respectively. Correlation of this data with Raman analysis of suspended and supported graphene samples implied a strong role of the substrate and irradiation sequence in determining the level of doping. A correspondingly higher reduction in mobility per incident electron was also observed for GFETs subjected to multi-step irradiation compared to single step, in line with measured Raman ID/IG ratios. Additionally, the Raman G-band DeltaFWHM variation was strongly dependent on the nature of the e-beam irradiation and the presence of the substrate. Single-step irradiated, substrate-supported graphene exhibited substantial broadening while multi-step irradiation resulted in G-band narrowing. This behavior was not observed for suspended graphene which indicated the addition or elimination of substrate-induced phonon-relaxation mechanisms in response to each type of irradiation. The narrowing of the FWHM (G) in the multi-step case is attributed to doping consistent with the Dirac point shift of ∼ -16V and the removal of Landau phonon damping above Ef > ℏwG2 . In strong contrast, single step irradiation of substrate supported graphene yielded a broadening of the FWHM (G) accompanied by a CNP shift of ∼ -8V indicating appreciable n-doping. This reveals the presence of alternate phonon decay channels even when Landau damping above Ef > ℏwG2 is removed. It is proposed in this dissertation that this phenomenon is linked to hybridization of silicon oxide defect states (induced by single-step e-beam irradiation) and graphene electron states. This hybridization promotes a graphene phonon decay channel distinct from Landau damping, the latter being forbidden under sufficient doping. It is proposed that the alternate phonon decay channel involves two-component inelastic scattering, wherein the graphene phonons transfer energy to the carriers in the lattice which in turn couple to the polar phonons of the substrate resulting in mobility reduction. Furthermore, it is proposed that this defect-induced, graphene phonon decay channel is inhibited in multi-step e-beam irradiation due to the presence of adsorbates on the graphene introduced during ambient exposure between radiation cycles. On e-beam irradiation the adsorbates induce polar orientation of water dipoles at the graphene/SiO2 interface. This polar layer shifts the hybridized defect bands closer to the graphene Dirac bands thereby reducing the inelastic scattering and inhibiting the phonon decay medicated by SiO2 surface polar phonons (SPP). This model also explains the enhancement of n-type doping in GFETS observed for multi-step irradiation. These results highlight the impact of substrate defects and interaction of induced defectivity with the e-beam along with the role of interfacial water in impacting graphene device performance.;The thesis also presents data on Raman-based characterization of graphene including layer number determination and carrier concentration measurement. Determination of layer number for graphene exfoliates focused on the splitting of the 2D Raman band. In addition, an alternate Raman-based thickness metrology was evaluated for CVD-based, polycrystalline graphene. Both were carried out on split gate test structures as a method for monolayer or bilayer confirmation in device geometries. In addition, carrier concentration measurements of exfoliates on 300nm SiO2 and split-gate test structure substrate have also been characterized with back gate biasing. These measurements made use of the stiffening of the Raman G-band with doping and the narrowing of the G-band FWHM. These results were important for validating conclusions from the e-beam irradiation experiments mentioned above regarding carrier doping.