A Simple Semiempirical Short-Channel MOSFET Current–Voltage Model Continuous Across All Regions of Operation and Employing Only Physical Parameters

Ali Khakifirooz,Osama M Nayfeh,Dimitri Antoniadis

doi:10.1109/ted.2009.2024022

Ali Khakifirooz, Osama M Nayfeh + Show 1 more

Open Access

https://doi.org/10.1109/ted.2009.2024022

Copy DOI

Abstract

A simple semiempirical model I D (V GS , V DS ) for short-channel MOSFETs applicable in all regions of device operation is presented. The model is based on the so-called ldquotop-of-the-barrier-transportrdquo model, and we refer to it as the ldquovirtual sourcerdquo (VS) model. The simplicity of the model comes from the fact that only ten parameters are used. Of these parameters, six are directly obtainable from standard device measurements: 1) gate capacitance in strong inversion conditions (typically at maximum voltage V GS = V dd ); 2) subthreshold swing; 3) drain-induced barrier lowering (DIBL) coefficient; 4) current in weak inversion (typically I off at V GS = 0 V) and at high V DS ; 5) total resistance at V DS = 0 V and V GS = V dd and 6), effective channel length. Three fitted physical parameters are as follows: 1) carrier low-field effective mobility; 2) parasitic source/drain resistance, 3) the saturation region carrier velocity at the so-called virtual source. Lastly, a constrained saturation-transition-region empirical parameter is also fitted. The modeled current versus voltage characteristics and their derivatives are continuous from weak to strong inversion and from the linear to saturation regimes of operation. Remarkable agreement with published state-of-the-art planar short-channel strained devices is demonstrated using physically meaningful values of the fitted physical parameters. Moreover, the model allows for good physical insight in device performance properties, such as extraction of the VSV, which is a parameter of critical technological importance that allows for continued MOSFET performance scaling. The simplicity of the model and the fact that it only uses physically meaningful parameters provides an easy way for technology benchmarking and performance projection.

Full Text