Local iterative Monte Carlo analysis of electron-electron interaction in short-channel Si-MOSFETs

The effects of electron-electron interaction on the electron distribution, substrate current, and gate current in short n-channel metal-oxide-semiconductor field-effect transistors (MOSFETs) are studied using the local iterative Monte Carlo (LIMC) approach. The complete distribution function is always available at each step of this approach and with reduced noise with respect to standard Monte Carlo (MC) simulation. Therefore, electron-electron interaction can be evaluated efficiently using scattering rates, allowing one to examine hot carrier effects that may play an important role for device reliability and characterization. Results for MOSFETs with channel length as short as 25 nm show that electron-electron interaction leads to an increase of the high energy tail of the electron distributions at the transition from channel to drain. The electron density around 3 eV is significantly increased even if the applied voltage is in the 1.0 V range.

I. Introduction

The Monte Carlo (MC) technique for the simulation of semiconductor devices is widely accepted as a tool to gain a detailed understanding of physical phenomena in modern semiconductor devices [1] [2] [3] [4] [5]. One advantage of MC approaches is that the full band structure can be readily included using numerical tables [6]. The simulation of carrier–carrier interaction is also possible but the short-range carrier–carrier scattering is particularly difficult to evaluate and computationally expensive [7]. A scattering rate approach normally treats the two-body problem, presenting the additional complication that a partner carrier must be considered to respect conservation laws. This can never be done exactly due to the limited sampling of particles available in a given simulation cell. A molecular dynamics model gives a complete description of the semiclassical many-body charge interaction in the short range, but it requires an expensive three-dimensional (3-D) approach [8], [9].