Deterministic simulation of electron and phonon transport in III-V devices

尖端的III-V半导体器件正在将电子产品推向更高的频率，这样它们就可以用于安全应用，例如汽车的近场雷达或高速数据传输。这些渡越和最大振荡频率超过300 GHz的极快器件不能再用经典技术的计算机辅助设计(TCAD)工具建模，需要更基本的基于Boltzmann方程的模拟方法。由于电子输运和由于自加热而降低器件性能的声子输运在这些器件中是重要的，因此对于这两种粒子，玻尔兹曼方程必须以自洽的方式求解。这需要一种数值方法，这超出了最先进的蒙特卡罗模拟器的可能性。最近开发了一种用于硅器件的确定性Boltzmann方程求解器，它具有TCAD所需的所有特征，并且可以模拟静态和小信号器件的特性和电子噪声。该模拟器将扩展到包括声子在内的III-V材料。该项目的主要重点将是开发模拟器的热部分，该部分将基于像电子部分一样的球谐展开。为了在设计过程中使用模拟器，它必须是数值稳定和高效的。新开发的模拟器将用于调查真实设备，测量数据将由合作伙伴提供，并探索III-V设备的最终限制。

Leading-edge III–V semiconductor devices are pushing electronics to higher frequencies such that they can be used, for example, in safety applications like near field radar for cars or in high speed data transmission. These extremely fast devices with transit and maximum oscillation frequencies beyond 300GHz can be no longer modeled with classical Technology Computer Aided Design (TCAD) tools and more fundamental simulation approaches based on the Boltzmann equation are required. Since electron transport and due to self-heating, which degrades the device performace, phonon transport are of importance in these devices, Boltzmann equations have to be solved for both types of particles in a self-consistent manner. This requires a numerical approach, which goes beyond the possibilities of the state-of-the-art Monte Carlo simulators. Recently a deterministic Boltzmann equation solver has been developed for silicon devices, which has all the required TCAD features and can simulate the stationary and small-signal device characteristics and electronic noise. This simulator will be extended to III-V materials including phonons. The main focus of the project will be on the development of the thermal part of the simulator, which will be based on a spherical harmonics expansion like the electron part. In order to use the simulator in a design process, it must be numerically stable and efficient. The newly developed simulator will be used to investigate real devices, for which the measurement data will be provided by a partner, and to explore the ultimate limits of III-V devices.