Quantitative identification of significant k-points enabling accurate & computationally efficient UTB MOS device simulation

Abstract

Ahstract-A semi-empirical $\mathbf{sp}^{3}\mathbf{d}^{5}\mathbf{s}^{*}$ tight-binding model is used to simulate the full-band structure, which in turn is used to calculate the electrostatics of Ultra- Thin-Body (UTB) MOS devices through a self-consistent solution with the 1-D Poisson's equation. Through the identification of significant k-points from the full band structure and calculation of electrostatics of UTB MOS devices over those k-points a computationally efficient yet accurate approach valid over a wide range of channel thicknesses, gate voltages and device temperatures has been shown. In this work, we focus on the practical implementation of this approach, where we show that the choice of a semi-empirical parameter η, which is a weighting factor (multiplier) to the maximum Fermi-Dirac probability (corresponding to the conduction band minima), is critical in identifying the significant k-points. Through an algorithm to quantify η, which is then used to accurately and effectively calculate device electrostatics, we obtain the optimal value of η, showing that this value, for a particular device temperature, is channel thickness independent while varying with gate voltage.