The aim of this work concerns the simulation of bulk semiconductors materials for the extraction of microscopic parameters as the drift velocity, the mean electron energy, and low-field electron mobility. Semiconductor under investigation are Silicon, GaAs and 4H-SiC.
Motivations regarding simulations and computational electronics are presented in the first part of the presented work. Descriptions on semiconductor bands structure and several scattering mechanisms are reported as well.

The non parabolic ellipsoidal band structure model is employed. For Silicon three equivalent X conduction bands are considered. The anisotropy of the conduction band is taken into account. The scattering processes taken into account are those due to the non-polar optical phonon, acoustic phonon and impurity scattering. For GaAs a two-valley model (G-valley and L-valley) is employed instead. The scattering mechanisms included are those due to polar optical phonon, non-polar optical phonon, acoustic phonon and impurity scattering. The band structure model used for 4H-SiC consist of a single equivalent non parabolic ellipsoidal M-valley, the anisotropy of the conduction band is taken into account. Electron scattering by polar optical phonon, non-polar optical phonon, acoustic phonon, ionized impurity scattering as well as the impact ionization have been included in our model. Both zero and first interactions have been taken into account for polar optical phonon. The Keldysh formula has been used for the impact ionization rate.

The analysis of the carrier transport in the semiconductors has been done using Monte Carlo method, description of the method is presented. Mainly the single-particle Monte Carlo method has been applied to simulate the motion of a single carrier in the momentum space. The ensemble Monte Carlo method has been employed for the analysis of transient carrier motions. The method is essentially dynamic.

All presented simulations were obtained for bulk materials that means without boundary conditions and were performed using Matlab 7. The results have been verified by making comparison with measurements and simulations available in literature. A program has been written in C++ using Borland Developer Studio 2006 allowing any user to perform the simulations for each semiconductor.