Model Order Reduction for Multiband Quantum Transport Simulations and its Application to p-Type Junctionless Transistors

Abstract

An efficient method is developed for multiband simulation of quantum transport in nanowire electronic devices within nonequilibrium Green's function formalism. The efficiency relies on a model order reduction technique, which projects the <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">k · p</i> Hamiltonian into a much smaller subspace constructed by sampling the Bloch modes of each cross-section layer. Several sampling approaches are discussed to obtain a minimum and accurate basis with reduced computational overhead. The technique is verified by calculating the valence bands of silicon nanowires (SiNWs) and by solving I-V curves of p-type SiNW transistors. This enables us to study for the first time the performances of large cross-section p-type junctionless (JL) transistors in the quantum ballistic transport limit. The influences of doping density, transport direction, channel length, and cross-section size are examined. We find that larger doping densities may lead to worse sub-threshold slopes due to the enhanced source-to-drain tunneling. Compared with their counterparts, i.e., classical inversion-mode (IM) transistors, they have better sub-threshold behaviors, but they do not necessarily provide a better ON/OFF ratio except when the channel is short or thin. In addition, unlike IM transistors, [110] and [111] channel directions in JL transistors are very robust against channel thicknes scaling.