(Color online) SWNTs on the Si(100) surface in two configurations: parallel over the dimer trench [(a) and (b)] and perpendicular to the Si(100) surface dimer trench [(c) and (d)]. is the distance from the bottom of the SWNT to the bottom of the Si(100) surface dimer trench obtained in the total energy minimization. The vertical displacements involved in the surface reconfiguration and , the angle of rotation of the SWNT about its axis, are also indicated. An evident “squashing” effect—an elongation along the direction—can be observed for the (9,3) nanotube but it was not present in any semiconducting nanotube. Note that for chiral tubes, a rotation about their axis and a displacement (not shown) along nanotube’s axis are linearly dependent. To minimize the cell size, a surface reconstruction was employed when placing semiconducting nanotubes, and the surface reconstruction for the (9,3) nanotube.
(Color online) Distance from the bottom of the dimer trench to the lowest carbon atom as a function of nanotube diameter for fixed . Semiconducting SWNTs parallel to the dimer trench will be closer to the underlying surface by about as compared to semiconducting SWNTs in the perpendicular configuration for this diameter range. Dashed lines are drawn as a guide to the eyes to facilitate the visualization of trends.
(Color online) The properties of the hybrid system do not depend on the underlying surface reconstruction as evidenced by the (9,3) nanotube on the reconstructed surface. The distance between the nanotube and surface is very close to that found for this nanotube on the reconstructed surface, see Fig. 2. Notice also the vertical elongation appearing again and the overall structural similarity to the structure presented in Fig. 1.
(Color online) Charge density isosurface plots to visualize the relative strengths of C–Si bonds. Values of the densities associated with each isosurface are shown. The strongest bond occurs for the metallic SWNT.
(Color online) Ab initio band structures and projected densities of states for SWNTs in different alignments with respect to the Si(100) surface. The first column shows the band structures and PDOS when SWNTs are aligned on top and parallel to the dimer trench, while the second column depicts band structures when the nanotubes are perpendicular to the dimer trench. Those results are obtained after atomic relaxation was performed. The ab initio results show a drastic reduction of the semiconducting gap for the hybrid system composed of semiconducting tubes and the Si(100) surface and a high degree of band hybridization. The different positions for the , points reflect the difference in size of the unit cells considered. The band structures involving the metallic nanotube (in blue) show no gap opening at the Fermi level. The band structures are calculated with a MP grid, while the PDOS was obtained with at least a MP grid. Arrows indicate the contribution to the band structure from carbon atoms in the vicinity of the Fermi level.
(Color online) Band structure for the (8,4) and (9,3) nanotubes around the Fermi energy to visualize gap openings or lack thereof. We omit zooming into band structures for which the gaps can be directly visualized from Fig. 5.
(Color online) Wave functions depicting -point states right below (highest occupied electronic state) and above (lowest unoccupied electronic state) the Fermi energy. The hydrogen bottom layer is not shown.
(Color online) Energy gain vs axial rotation, prior to the relaxation cycle and at fixed height. Red squares show results in the parallel configuration, while the blue triangles correspond to the perpendicular configuration. . The results shown here help identify the best angular configuration for chiral nanotubes, which can not be known a priori. The largest adsorption energies occur for tubes parallel to the trench. The energy dependence on angle of rotation is more marked as the nanotube diameter is decreased, since it is brought in closer proximity with the surface. The metallic nanotube shows the most pronounced energy dependence.
Lattice constants obtained with our bases, to exemplify the flexibility of our basis set (in Å).
Carbon nanotubes with length in their unit cell commensurate within 2% to the Si(100) surface supercell. In bold, the respective values for the (6,6) SWNT previously studied by Orellana et al. (Refs. 3 and 5).
Averaged displacement (over the unit cell) in the direction (Å) of silicon atoms closest to the nanotube; c.f. Fig. 1. refer to the parallel configuration and to the tubes in the perpendicular configuration.
Charge density at the bonds between nanotube carbon and surface silicon atoms. Notice the smaller density for all studied semiconducting SWNTs in comparison with the bond densities for the metallic nanotube.
Amount of electronic charge per unit length transferred from the SWNT to the slab from the Voronoi (Hirshfield) deformation density analysis.
Electronic gaps (in eV) for the hybrid systems composed of semiconducting nanotubes on the Si(100) surface. We also include the (9,3) tube for completeness.
Adsorption energies per unit length (eV/Å) after full relaxation has been achieved. Notice the lower adsorption energies obtained for semiconducting nanotubes. For the (12,4) tube (in bold) the perpendicular configuration turns out to be the most favorable.
Article metrics loading...
Full text loading...