Study of Electron Transport in SOI MOSFETs Using Monte Carlo Technique with Full-Band Modeling

Effects of conduction-band non-parabolicity on electron transport properties in silicon-on-insulator (SOI) metal-oxide-semiconductor field effect transistors (MOSFETs) are studied by performing Monte Carlo simulation with a full-band modeling. An empirical pseudo-potential method is adopted for evaluating the two-dimensional electronic states in SOI MOSFETs. SOI-film thickness dependence of phonon-limited mobility, drift-velocity and subband occupancy is calculated and the results are compared with those of a simple effective mass approximation. The non-parabolicity effects are found to play an important role in 4-fold valleys under higher applied electric fields or at higher temperatures.     

Wave-mixing effects on electronic noise in semiconductors

The results of a Monte Carlo analysis of hot-electron intrinsic noise in a n-type GaAs bulk driven by two mixed large-amplitude alternating electric fields having frequency in the subterahertz range are presented. The noise properties are investigated by studying the velocity autocorrelation function and the noise spectrum. We explored the relations among the frequency response and the velocity fluctuations as a function of the frequencies and intensities of the mixed fields. When the semiconductor is driven by two mixed ciclostationary electric fields, a resonant-like enhancement of the spectra near the two frequencies of the applied fields is found.     

On the Electron Transient Response in a 50 nm MOSFET by Ensemble Monte Carlo Simulation in Presence of the Smoothed Potential Algorithm

The inclusion of a smoothed potential algorithm within the Ensemble Monte Carlo method (EMC) to account for quantization effects in the inversion layer of a silicon n-MOSFET has been discussed by several authors. Most of the data reported deal with steady state terminal current, transconductance, and capacitance. Within this approach, the electric field acting on each particle is computed from the smoothed potential, which introduces a potential barrier underneath the gate region that pushes the carriers away from the interface, thus accounting for space quantization effects. However, in the EMC method, the electric field at the interface is also used to compute the displacement charge/current during the transient regime. In the implementation of the smoothed potential algorithm, care must be taken when computing this component of the total gate charge. We distinguish between two differently computed electric fields, one from the smoothed potential used for the charge transport and the other one computed from the real potential, as obtained from the solution of Poisson's equation, and used for the displacement charge. We propose this procedure in order to properly include space quantization effects, and at the same time avoid the inaccuracy introduced by the smoothed potential in the displacement charge.     

Sensitivity of single- and double-gate MOS architectures to residual discrete dopant distribution in the channel

The effect of a single discrete impurity in the channel of Fully-Depleted Single- and Double-Gate MOSFETs is analyzed by means of 3D Monte Carlo simulation. The Double-Gate (DG) architecture appears to be less sensitive to the dopant perturbation than the Single-Gate (SG) counterpart. For an N-channel device the influence of a P-type impurity on the current-voltage characteristics is shown to be strongly dependent on the impurity position in the channel. The maximum current degradation is obtained for an impurity located about 5 nm from the source-end of the channel. The I _on reduction reaches 6% in DG and 10.5% in SG. A small current enhancement (less than 2%) is induced by an N-type impurity. These results are analyzed in terms of velocity profile between source and drain.     

A many-particle quantum-trajectory approach for modeling electron transport and its correlations in nanoscale devices

Electron transport in mesoscopic systems is analyzed in terms of quantum (Bohm) trajectories associated to wave-function solutions of a many-particle (effective-mass) Schrödinger equation. Many-particle Bohm trajectories can be computed from single-particle Schrödinger equations. As an example, electron correlations for a triple-barrier tunneling system with electron-electron interactions are computed. Simulated noise results for interacting electrons that tunnels through triple barriers are presented. The approach opens a new path for studying electron transport and quantum noise in nanoscale systems, beyond the “Fermi liquid” paradigm.     

Full-Band Monte Carlo Simulation of Two-Dimensional Electron Gas in SOI MOSFETs

A full-band Monte Carlo simulation of two-dimensional electron gas is performed to study effects of the non-parabolicity of the energy band structure on the phonon-limited electron mobility in SOI MOSFETs with a thin Si-layer.     

3-D Parallel Monte Carlo Simulation of Sub-0.1 Micron MOSFETs on a Cluster Based Supercomputer

A full band, three-dimensional, Monte Carlo simulator for deep sub-micron Si MOSFET like devices has been developed, with the goal to obtain optimal performance on a parallel system built from a cluster of commodity computers. A short-range carrier-carrier and carrier-ion model has been implemented within this framework, using Particle-Particle Particle-Mesh (P3M) algorithm. Test simulations include the 90 nm well-tempered MOSFET for which measurements are available. Simulation benchmarks have identified several factors limiting the overall performance of the code and suggestions for improvements in these areas are made.     

Tetrahedral elements in self-consistent parallel 3D Monte Carlo simulations of MOSFETs

Novel thin-body architectures with complex geometry are becoming of large interest because they are expected to deliver the ITRS prescribed on-current when semiconductor transistors are scaled into nanometer dimensions. We report on the development of a 3D parallel Monte Carlo simulator coupled to a finite element solver for the Poisson equation in order to correctly describe the complex domains of advanced FinFET transistors. We study issues such as charge assignment, field calculation, treatment of contacts and parallelisation approach which have to be taken into account when using tetrahedral elements. The applicability of the simulator is demonstrated by modelling a 10 nm gate length double gate MOSFET with a body thickness of 6.1 nm.     

3D Monte Carlo Modeling of Thin SOI MOSFETs Including the Effective Potential and Random Dopant Distribution

We use the effective potential to include quantum mechanical effects in thin SOI MOSFETs simulated with 3D Monte Carlo. We explore the role of discrete dopant distributions on the threshold voltage of the device within the framework of the effective potential by examining the current-voltage behavior as well as the electron distributions within the device. We find that simulations with the effective potential produce a similar shift in current as classical simulations when the dopants are considered to have a random discrete distribution instead of a uniform distribution.     

3D Monte Carlo Analysis of Discrete Dopant Effects on Electron noise in Si Devices

A preliminary study of velocity fluctuations in simple devices using an atomistic model for the ionized impurity scattering is presented. The velocity fluctuations are responsible for thermal noise in semiconductor devices. In this paper the influence on the spectral density of velocity fluctuations of the boundary conditions, the doping level, the length of the resistors and the excess/default of impurities has been addressed.     

Band-to-Band Tunneling by Monte Carlo Simulation for Prediction of MOSFET Gate-Induced Drain Leakage Current

Gate induced drain leakage (GIDL) current caused by band-to-band tunneling is studied by Monte Carlo simulation with ballistic least-action trajectory integration. Together with weak inversion and early subthreshold simulation by drift-diffusion formalism, the entire range of the OFF-state drain current can be predicted for technology evaluation. The methodology is demonstrated by a case study for source/drain asymmetry super-halo design.     

Full-Band Particle-Based Analysis of Device Scaling for 3D Tri-Gate FETs

In this work, a 25 nm gate length three-dimensional tri-gate SOI FET with a wrap around gate geometry is studied using a full-band particle-based simulation tool. The tri-gate FETs have shown superior scalability over planar device structures, reduction of short channel effects, higher drive currents and excellent gate-channel controllability compared to their planar counterparts. Simulations were performed by scaling the length and the width of the tri-gate SOI FET channel to study its short-channel and short-width effects. The influence of the scaling on the dynamic response has also been explored by performing a frequency analysis on the device.     

Boundary conditions for Density Gradient corrections in 3D Monte Carlo simulations

Monte Carlo remains an effective simulations methodology for the study of MOSFET devices well into the decananometre regime as it captures non-equilibrium and quasi-ballistic transport. The inclusion of quantum corrections further extends the usefulness of this technique without adding significant computational cost. In this paper we examine the impact of boundary conditions at the Ohmic contacts when Density Gradient based quantum corrections are implemented in a 3D Monte Carlo simulator. We show that Neumann boundary conditions lead to more stable and physically correct simulation results compared to the traditional use of Dirichlet boundary conditions.     

The Effective Conduction Band Edge Method of Quantum Correction to the Monte Carlo Device Simulation

To include quantum effects, a quantum correction is made to the semi-classical Monte Carlo (MC) simulation by the effective conduction band edge (ECBE) method. The quantum corrected potential energy can be calculated from the classical potential energy by the ECBE equation and thus the quantum mechanical force in the simulation replaces the classical force. Under the non-equilibrium condition, carriers have a temperature different from the lattice. For the simulation of a double-gate MOSFET, we replace thermal energy in the ECBE equation with the average value of the stress tensor along each transverse line, to account for the variation of the electron “temperature” along the longitudinal direction. A 3 nm thick double gate nMOSFET is simulated. The result shows that electrons now see a higher barrier from the source to the drain if the carrier temperature is considered, resulting in a smaller drain current compared to that obtained from the previous ECBE method.     

基于随机统计分析的黑启动操作过电压的计算校验

针对电力系统黑启动过程中空充输电线路引起的过电压问题,提出利用蒙特卡洛法建立用于电力系统操作过电压计算的概率模型.通过随机抽样给出开关操作信息,结合贝瑞隆输电线路数学模型,开发了用于黑启动过程中过电压校验的计算与统计分析程序,实际算例表明了本文中模型、算法和程序的有效性.     

Using probability distribution functions in reliability analyses

The use of probability distributions to describe reliability inputs such as failure rates and restoration times is fairly new compared to using average values for parameters. Predictive indices of composite power system reliability are frequently calculated but not directly used in the decision-making process, possibly because the indices are usually described using average values and convey little information about risk. More information can be derived and a different impression given by indices of a reliability analysis when probability distributions are used to describe the inputs and outputs, as shown in examples presented in this paper.     

Introducing energy broadening in semiclassical Monte Carlo simulations

Combining an insight on the quantum transport given by the Wigner function formalism and the classical perturbation theory, an algorithm has been developed that allows the introduction of collisional broadening in semiclassical electron transport Monte Carlo (MC) simulations. In the proposed algorithm, electron energy and momentum are treated as independent variables; the laws of energy and momentum conservation are fulfilled at each scattering event, but the relationship between energy and momentum is not given by the traditional expression, since Bloch states are not eigenstates of the total Hamiltonian. The results obtained for a simple model semiconductor demonstrate that the non-physical instabilities observed in previous attempts to introduce collisional broadening in semiclassical MC simulations have been removed. The algorithm is suitable for application in MC simulations of realistic device models.