A TCAD-ready, analytical electron mobility model for strained bulk Si has been developed. The low-field mobility model includes doping dependence, temperature dependence, substrate orientation dependence and dependence on the field direction. Electron velocity characteristics at high electric fields have been calculated using a full-band Monte Carlo code. A widely used high-field mobility model has been augmented for strain dependence. All parameters depend on the valley splitting via simple, analytical expressions. The model is directly fit to the Monte Carlo data. Electron transport in the channels of single-gate and ultra-thin body double-gate devices has been studied using a sub-band Monte Carlo code. Degenerate statistics, (100) and (110) substrate orientations, and strain effects have been taken into account. Effects of degeneracy and of inter-sub-band scattering were found to be very important to explain measurement data. The development of a Wigner Monte Carlo code for the simulation of silicon FETs near the scaling limit has been continued. A self-consistent iteration loop has been implemented. A particle annihilation algorithm needed to cope with the negative sign problem has been improved, yielding increased robustness and efficiency of a Monte Carlo simulation. VSP is a multi-purpose Schroedinger Poisson solver for TCAD applications. For the investigation of gate stacks, models are included for interface traps and bulk traps. For leakage current calculation, carriers tunneling from quasi-bound states and from free states are taken into account. Direct and trap-assisted tunneling current components are considered. Ballistic quantum transport through one-dimensional mesoscopic structures is solved numerically using the quantum transmitting boundary method. For dissipative quantum transport in carbon nanotube FETs, the non-equilibrium Green's function method (NEGF) method is employed. The effects of elastic and inelastic phonon scattering on device performance have been analyzed in detail. DC and AC characteristics of CNT-FETs have been optimized by variation of geometry parameters such as source-gate and drain-gate spacer widths. A kinetic Monte Carlo simulator has been implemented to investigate electric currents in organic semiconductor devices based on ordered and disordered molecular solids. The density-of-states model in use is based on the Gaussian disorder model. The simulator has been applied to doped and undoped zinc pthalocyanine (ZnPc) films, for which current-voltage characteristics have been achieved for film-lengths up to 140 nm. The simulator includes charge injection and ejection models at the contacts, which is important since the current-enhancing effect of molecular doping essentially occurs at the device's contact-interfaces. Charge transport in organic semiconductors has also been investigated theoretically. For the different transport regimes, different approaches have been pursued. A mobility model has been formulated using the variable range hopping conduction mechanism in the presence of an electric field. This model describes both the electric field and temperature dependencies of mobility. The mobility model has been extended to consider small-polaron hopping transport. An analytical model for electrical conductivity based on percolation theory has been derived. In disordered organic semiconductors and polymers, conductivity proceeds via hopping between localized states, being embedded into a random medium. Dopant atoms or molecules are subjected to positional and energy disorder. Therefore, a double exponential function should be a realistic model for density of states (DOS) distribution. Using the new DOS function, a model of hopping charge carrier transport in doped disordered organic semiconductors has been obtained. This model can explain recent experimental data and results of Monte Carlo simulations. Traps strongly affect the charge transport since trapped carriers no longer take part in charge transport.
|