Search results

The Monte‐Carlo particle model is a technique of simulating small semiconductor devices. It consists briefly of following the detailed transport histories of individual carriers, their time of free flight and consequent scattering chosen by a random number technique. A description of the method is given. The method has proved itself successful in semiconductor analysis, and as an example of its application we are using it to study the influence the epitaxial doping has on the performance of field‐effect transistors. We are comparing a transistor with an epitaxially grown active layer, with one with an ion implanted active layer and with an ideal device with an abrupt transition between the epilayer and the substrate. The cut‐off bias for ideal transistor is found to be more sharply defined than for the other two types of transistors. The spatial distribution of the carriers follows roughly the doping profile near the source. Underneath the gate the peak of the carrier density is pushed further down and into the substrate as the gate bias increases. This peak also weakens as the gate bias rises, and vanishes at, and beyond cut‐off. In the high field region after the gate the upper valleys population increases with increased drain bias and decreases with increased gate bias. The power gain and the y‐parameters are examined for all devices, both near pinch‐off and for no external gate bias. In both cases the ion implanted transistor shows the greatest gain. This transistor also exhibits the lowest minimum noise figure.

The response of a MESFET and an inverted HEMT to the impact of an a particle has been calculated by means of the Monte Carlo Particle Model, a technique for solving Boltzmann's transport and Poisson's field equation self‐consistently in space and time. The calculations show that all the terminals of the MESFET react by generating an initial current pulse followed by another; the timing of the second pulse depends on the angle of incidence of the α particle. The lattice heating rate was found to be largest at the corners of the Ohmic contacts. The HEMT, on the other hand, hardly reacts electrically to the α particle but is more likely to burn out in an a particle radiation environment because of the larger lattice heat generation taking place in the interior of the transistor. The results also support the theory of the hot‐electron induced subsurface catastrophic failure mechanism.

An empirical velocity‐field relationship, based on Monte Carlo simulation, is used to modify a drift‐diffusion model for the characterization of short gate GaAs MESFET's. The modified drift‐diffusion model is used to generate both the steady‐state and the small‐signal parameters of submicron GaAs MESFET's. The current, transconductance, and cutoff frequency are compared with two‐dimensional Monte Carlo simulation results on a 0.2 µm gate‐length. The model is also used to predict measured I‐V and s‐parameters of a 0.5 µm gate‐length ion‐implanted GaAs MESFET. The comparison and the analysis made, support the accuracy of the modified drift‐diffusion simulator and makes it computationally efficient for analysis of short‐gate devices.

A quasi‐two‐dimensional analytical model for GaAs MESFETs is proposed. It enables the calculation of the dc, the small‐signal, and the noise behaviour of GaAs MESFETs and takes into account both doping and low‐field mobility profiles in the active layer of the transistor. It is shown that the profile of the low‐field mobility near the bottom of the active layer has a considerable influence on the minimum noise figure.

The purpose of this paper is to set up a consistent off‐equilibrium thermodynamic theory to deal with the self‐heating of electronic nano‐devices.

From the Bloch‐Boltzmann‐Peierls kinetic equations for the coupled system formed by electrons and phonons, an extended hydrodynamic model (HM) has been obtained on the basis of the maximum entropy principle. An electrothermal Monte Carlo (ETMC) simulator has been developed to check the above thermodynamic model.

A 1D n⁺−n−n⁺ silicon diode has been simulated by using the extended HM and the ETMC simulator, confirming the general behaviour.

The paper's analysis is limited to the 1D case. Future researches will also consider 2D realistic devices.

The non‐equilibrium character of electrons and phonons has been taken into account. In previous works, this methodology was used only for equilibrium phonons.

The dynamical characteristics of electron transfer between two channels are elucidated by using a many‐particle Monte Carlo model with self‐consistent electric fields. The study has been performed to assess switching speeds associated with various novel devices such as velocity modulation transistors and dual channel high electron mobility transistors. Typical time constants for a one micrometer device (0.4 µm gate length) are 3.5 psec for the longitudinal (source‐to‐drain) and 0.2 psec for the transport perpendicular to the interfaces between the two channels.

By generalizing the equation of state of the conduction electron gas in a semiconductor to include a dependence not only on electron density but also on the density gradient we show that the standard diffusion‐drift description can be extended to describe much of the quantum mechanical behavior exhibited by strong inversion layers.

The purpose of this paper is to deal with the self-heating of semiconductor nano-devices.

Transport in silicon semiconductor devices can be described using the Drift-Diffusion model, and Direct Simulation Monte Carlo (MC) of the Boltzmann Transport Equation.

A new estimator of the heat generation rate to be used in MC simulations has been found.

The new estimator for the heat generation rate has better approximation properties due to reduced statistical fluctuations.

A new algorithm for the inhomogeneous Boltzmann equation in one spatial and velocity dimension, based on the method of characteristics, is presented. Using the analytic solution to the Boltzmann equation along its characteristics, the solution to the classical transport problem in semiconducting structures within the relaxation time approximation for the collision integral is obtained iteratively. An n+nn+‐diode is studied numerically and the effects of ballastic electrons on low‐order moments are investigated.

A one‐dimensional finite difference scheme adapted to high order moment equation models arising in the approximate description of semiconducting submicron structures is presented. The new scheme is a natural extension of the Scharfetter‐Gummel scheme used in drift‐diffusion models. Through local analytic solutions an accurate representation of exponentially varying solution components is realised.