Silicon nanowire-based complex structures : A Large-scale atomistic electronic structure and ballistic transport

While the hierarchical assembling as well as the dramatic miniaturization of Si nanowires (NWs) are on-going, an understanding of the underlying physics is of great importance to enable custom design of nanostructures tailored to specific functionalities. This work presents a large-scale atomistic insight into the electronic properties of NW-based complex structures, starting from the subsystem level up to the full assembly, within the framework of pseudopotential-based linear combination of bulk bands method. Laying the groundwork by grasping single Si NWs, we get into a large extent an unexplored territory of NW networks and kinked NWs. As one end product, a versatile estimator is introduced for the band gap and band-edge lineups of multiply-crossing Si NWs that is valid for various diameters, number of crossings, and NW alignments. Aiming for an exploration of the low-lying energy landscape, real space wave function analysis is undertaken for tens of states around band edges which reveal underlying features for a variety of crossings. Predominantly, the valence states spread throughout the network, in contrast the conduction minima are largely localized at the crossings. Given the fact that substantial portion of the band edge shift drives from the confined conduction states, branched Si NWs and nanocrystals have quite close band gap values as the networks of similar wire diameters. Further support to wave function analysis is provided via quantum ballistic transport calculations employing the Kubo-Greenwood formalism. The intriguing localization behaviors are identified, springing mainly at the crossings and kinks of NWs. The ballistic transport edge set apart the conducting extended states from the localized-band gap determining ones. Our findings put forward useful information to realize functionality encoded synthesis of NW-based complex structures, both in the bottom-up and top-down fabrication paradigms.