Vertically Grown Carbon Nanostructure Alignment: An Investigation by Using X-Ray Absorption Spectroscopy

International audienceAbstractX-Ray Absorption Spectroscopy (XAS) on the carbon K edge of carbon nanostructures (nanotubes, nanofibers, nanowalls) is reported here. They are grown on plain SiO2 (8 nm thick) /Si (100) sub - strates by a Plasma and Hot Filaments - enhanced Catalytic Chemical Vapor Deposition (PE HF CCVD) process. The XAS spectra are highly sensitive to a previous thermal treatment, as an intense adsorption on the outer wall of the nanotubes may strongly affect the absorption transitions. The morphology and the nature of these carbon nanostructures are characterized by SEM, TEM and Raman spectroscopy. According to conditions of catalyst preparation and DC HF CCVD process, carbon nanotubes (CNTs), carbon nanofibers (CNFs), carbon nanowalls (CNWs), carbon nanoparticles (CNPs) with different orientation of the graphene plans or shells can be prepared. From the angular dependence of the incident light and geometrical morphology of the nanostructures, wide variations of the C K - edge intensity of the transitions to the empty π* and σ* states occur. A full lineshape analysis of the XAS spectra has been carried out using a home -made soft - ware, allowing estimating the relative proportion of π* and σ* transitions. A geometrical model of the angular dependence with the incidence angle of the light and the morphology of the carbon nanostructures is derived. With normalization to the HOPG (Highly Oriented Pyrolytic Graphite) reference case, a degree of alignment can be extracted which is representative of the localized orientation of the graphitic carbon π bonds, accounting not only for the overall orientation, but also for local defects like impurities incorporation, structural defects,… This degree of alignment shows good agreement with SEM observations. Thus CNTs films display degrees of alignment around 50%, depending on the occurrence of defects in the course of the growth, whereas no special alignment can be detected with CNFs and CNPs, and a weak one (about 20%) is detected on CNWs