Numerical Simulation of Semiconductor Devices and Circuits for THz Applications

总体目标是开发一个用于检测和产生THz信号的电路中嵌入的半导体器件的仿真框架，并探索新的器件和电路概念。商业TCAD套件所基于的常用半导体方程将通过包含加速项扩展到THz信号的情况，这将使我们能够模拟2D和3D器件中的等离子体波。此外，我们将超越漂移-扩散近似(例如，能量传输或流体动力学模型)，因为这种近似对于高迁移率和短沟道是不适用的。必要的传输和噪声参数将通过基于玻尔兹曼传输方程的微观模拟方法产生，该方法可在ithe上获得。这些工具还将用于在同类条件下和在设备的情况下评估较简单的运输模型的准确性。半导体方程的变化将降低它们的数值性质和新的稳定概念，而不是Scharfetter-Gummel稳定化方案，其中必须考虑半导体方程的特殊限制(例如，负粒子密度在物理上是不可能的)。由于用于检测或产生THz信号的THz电路的基本工作原理是基于非线性效应的，因此需要一种大信号方法。所得到的半导体器件的离散方程系统将以紧凑的形式表示，因为太赫兹电路的分析需要大量的模拟来优化器件和周围电路。由于非线性，每个设置都需要完整的大信号分析。在大信号分析的情况下，我们将应用基于样条子波的方法与多速率偏微分方程技术相结合。在前人工作的基础上，发展了可处理大系统规模的振荡频率估计技术。此外，我们还将实现噪声计算的基本方法。在新开发的模拟框架下，我们将研究THz探测器和发生器电路的关键品质因数。例如，我们将调查包括所有寄生的标准硅MOSFET(体硅或FDSOI)在THz探测器电路中的性能，以及偏置条件和器件端口的端接如何影响响应度。此外，我们将评估不同的装置概念对太赫兹产生的适用性。我们将研究在现实条件下(例如，室温)是否有可能通过等离子体波产生太赫兹辐射。由于目前提出了大量新的器件概念，我们将灵活地将我们的研究适应于最有前途的器件。

The overall goal is to develop a simulation framework for semiconductor devices embedded in circuits for detection and generation of THz signals and to explore new device and circuit concepts.The usual semiconductor equations, on which the commercial TCAD suites are based, will be extended to the case of THz signals by inclusion of the acceleration term, which will enable the simulation of plasma waves in 2D and 3D devices.In addition, we will go beyond the drift-diffusion approximation (e.g. energy transport or hydrodynamic models), because this approximation fails for high mobilities and short channels. The necessary transport and noise parameters will be generated by microscopic simulation methods based on the Boltzmann transportequation, which are available at the ITHE. These tools will be also used to assess the accuracy of the simpler transport models under homogeneous conditions and in the case of devices. The changes in the semiconductor equations will degrade their numerical properties and new stabilization concepts beyondthe Scharfetter-Gummel stabilization scheme have to be developed, where the special restrictions of the semiconductor equations have to be considered (e.g. negative particle densities are physically impossible).Since the fundamental operating principles of THz circuits for detection or generation of THz signals are based on nonlinear effects, a large-signal approach is required. The resultant discrete system of equations for the semiconductor devices will be formulated in a compact form, because the analysis of THz circuits requires a huge amount of simulations for optimizing both the device and the surrounding circuitry. Due to the nonlinearity, each setting requires a full large-signal analysis. In the case of a large signal analysis, we will apply the spline-wavelet-based approach in combination with the multi-rate PDE technique. Techniques for estimating the oscillation frequency are to be developed which canhandle huge system sizes, based on the proposer's former work. In addition, we will implement basic methods for noise calculation.With the newly developed simulation framework, we will investigate key figures of merit of THz detector and generator circuits.For example, we will investigate how standard silicon MOSFETs (bulk or FDSOI) including all parasitics perform in THz detector circuits and how bias conditions and the termination of the device ports influence the responsivity. Furthermore, we will evaluate the suitability of different device concepts for THz generation. We will investigate whether it is possible to generate THz radiation by plasma waves under realistic conditions (e.g. room temperature). Since currently a large number of novel device concepts are suggested, we will flexibly adapt our research to the most promising ones.