A comparative study for quantum transport calculations of nanosized field-effect transistors

This article presents a comparative study on quantum mechanical frameworks between the widely used local Quasi-Fermi Level (QFL) model and a recently developed top of the barrier splitting (TBS) model. Both models are based on an atomistic quantum mechanical solver using the linear combination of bulk band method (LCBB). The QFL model uses the local Quasi-Fermi Level to represent the local equilibrium and calculate the occupied charge density as well as the current flow along the channel. The TBS model extracts scattering state information from the stationary solution of the system, then calculates the charge density as well as the ballistic and tunneling current. Using these two models, the 10 nm and 22 nm double-gate ultra-thin-body structures are simulated. Comparisons in occupied charge densities, self-consistent potentials as well as the I–V characteristics between these two models are presented. It is found that the QFL model significantly overestimate the subthreshold charge density inside the channel, as well as the current, while it works fine in the ON state of the device. It is also found that the results from both QFL and TBS models tend to coincide with each other as the drain bias approaching zero.

► We examine the validity of two quantum simulation methods named QFL and TBS.
► QFL overestimates subthreshold charge density and current at large drain bias.
► QFL works fine for ON states even under high drain bias.
► Results from TBS and QFL tend to coincide at small drain bias.
► Strict transport should be studied by scattering calculation or TBS approximation.