Non-stationary transport effects: impact on 0.1 μm PD SOI technology

When devices are scaled-down in the sub-0.1 μm domain, the classical drift-diffusion (DD) model fails to predict velocity overshoot and carrier diffusion due to electronic temperature gradients. Moreover, this model neglects the dependence of impact ionization on carrier energy, and thus advanced models become mandatory for accurate simulation of current devices. This is especially important for partially depleted (PD) SOI devices. Solutions such as Monte-Carlo (MC) simulation are difficult to apply for technology optimization (Fischetti and Laux, 1988) due to the CPU requirements, and therefore we used a modified energy balance (EB) model available in commercial tools. In this paper, we analyze how nonstationary effects affect PD SOI device behavior and we estimate quantitatively the real impact of technology (gate length, channel and/or LDD doping) on the level needed for carrier transport modeling (DD versus EB)