A microscopically accurate model of partially ballistic nanoMOSFETs in saturation based on channel backscattering

TL;DR

This paper introduces a more accurate model for partially ballistic nanoMOSFETs that improves upon the Lundstrom model by better representing energy distribution and inelastic scattering, validated through Monte Carlo simulations.

Contribution

A new model for channel backscattering in nanoMOSFETs that does not rely on the virtual source concept and aligns closely with Boltzmann transport equation results.

Findings

The model accurately estimates backscattering coefficients.

It outperforms the Lundstrom model in predicting energy distribution.

Validated by 2D Monte Carlo simulations.

Abstract

We propose a model for partially ballistic MOSFETs and for channel backscattering that is alternative to the well known Lundstrom model and is more accurate from the point of view of the actual energy distribution of carriers. The key point is that we do not use the concept of "virtual source". Our model differs from the Lundstrom model in two assumptions: i) the reflection coefficients from the top of the energy barrier to the drain and from top of the barrier to the source are approximately equal (whereas in the Lundstrom model the latter is zero), and ii) inelastic scattering is assumed through a ratio of the average velocity of forward-going carriers to that of backward-going carriers at the top of the barrier kv > 1 (=1 in the Lundstrom model). We support our assumptions with 2D full band Monte Carlo (MC) simulations including quantum corrections in nMOSFETs. We show that our model allows to extract from the electrical characteristics a backscattering coefficient very close to that obtained from the solution of the Boltzmann transport equation, whereas the Lundstrom model overestimates backscattering by up to 40%.