Study of atomic silicon dangling bonds for beyond-CMOS nanocomputing

The need for less power-consuming and more performing devices pushed CMOS engineering up to its limits, thus new methods of performing computing and creating larger memories and faster devices are becoming a necessity as the technologies that carry this promise (More-Moore, More than Moore or Beyond CMOS). The mechanism of interaction exploited by these technologies is different from the traditional voltage-to-current relation of CMOS, carrying on collateral that cannot be addressed in a standard way because of its complexity. Field-coupled nanocomputing emerges as a candidate in this landscape, with promisingly low power consumption, high-speed operation, and the intrinsic very compact size. The possible vectors used to transport information can be a variety thus a way of predicting with good accuracy and avoiding expensive ab initio simulations' computational costs has become more urgent. This thesis exploits the adaptability of MoSQuiTo methodology, seeking to determine if it is feasible to simulate atomic components to design field-coupled circuits from atomic Silicon Dangling Bonds (SiDBs) exploiting a molecular self-consistent potential algorithm. The first phase involved the definition of the geometry, electrostatic properties, and physical behavior under an annealing schedule of SiDB-like molecules, represented by two dots interacting through a screened Coulomb potential. In this first phase, the molecule definition and post-processing phases of MoSQuiTo have been gathered together to derive the figure of merits needed for the system-level analysis. In the second stage, a self-consistent method called SCERPA (Self-Consistent Electrostatic Potential Algorithm) is employed to simulate the information transmission in the circuit after its layouts have been design. The results can then be subjected to post-processing, allowing the extraction of the most relevant physical parameters like electrostatic potential and electric field in every point within the simulation space or the charge in each molecule, thus a full acknowledgment of the propagation of the information in the circuit.