Pearson versus gaussian effective potentials for quantum-corrected Monte-Carlo simulation

摘要

We propose a new effective potential for including quantization effects in nanometer scaled nMOSFET. It is calculated as the convolution of a Pearson IV distribution by the Poisson potential resulting from the 2D Poisson equation. Compared to the usual Gaussian distribution, the Pearson IV distribution, calibrated to fit the squared modulus of the ground sub-band Schroedinger's wave function, drastically improves the wave-packet description close to the oxide barrier. This approach has been implemented into a semi-classical Monte-Carlo particle simulator and tested in DGMOS capacitors. We find that the new Pearson Effective Potential gives a very good representation of the electron density profile inside the silicon film, which overcomes the well-known weakness of the Gaussian Effective Potential.