次世代パワー半導体用高精度コンパクトモデルの開発とスケーリング則の確立

摘要

Power electronics is one of the key technologies to realize a low emission society. Semiconductor power devices, which are the main components of power electronics instruments, are expected to improve in power loss. A lot of research and development have been done especially for IGBTs (Insulated Gate Bipolar Transistor), which are widely used for over 600V class rated voltage, such as EV (Electric Vehicle) and HEV (Hybrid Electric Vehicle). The annual demand for IGBTs is around 2.5 billion dollars annually in 2012, accounting for about 8% of the semiconductor market for Power Management. Its application range has rapidly expanded, from IH (Induction Heating) cookers, inverter air conditioners, and also high voltage DC transmission (HVDC) devices on behalf of several hundred megawatts of Thyristor. Its market is growing at an annual rate of more than 10%. Devices using new materials such as Gallium Nitride and Silicon Carbide as the next generation power devices have been prototyped. But Silicon devices, including the Silicon IGBTs are considered to be the mainstream for the next few decades due to its material cost and ease of processing. In the development of IGBT application systems, such as inverters, circuit simulation is used to improve its electrical efficiency and estimate heat loss and noise. A mathematically formulated model called the compact model that reproduces the characteristics of the IGBT, is used in the circuit simulation. But current IGBT compact models do not achieve enough accuracy, especially for latest trench-gate IGBT. In this study, accurate compact model development for the trench gate IGBT was achieved for the first time in the world by analyzing in detail the principle operation of the IGBT. The compact model represents high precision characteristics of the device in the simple formula and device structure parameters. This model has been published as a Quasi-2D MOS-ADE model. In addition, in the analysis of the compact model, the newly discovered Scaling Principle as well as CMOS technology was also present in the power devices. The possibility to realize the miniaturization of next-generation power devices was demonstrated. In Chapter 1 and Chapter 2, the background and purpose of the study is reviewed. Discuss on the detail the operating mechanism of the IGBT, in particular, the preparation for the construction of the model starts from Chapter 3. In Chapter 3, a new mathematical expression for the trench-gate IGBT based on the operation of the semiconductor device physics, was established as a compact model. This model makes possible the representation of potential and carrier distribution of the Cathode side with high accuracy. In this model, the current flows are assumed to be separated into three portions, the electron current and the hole current flowing through the mesa region that is sandwiched between the trenches, and the electron current flowing through the accumulation layer of the gate sidewalls where the current in the device has been represented as an equivalent circuit conventionally. As a result, the distribution of electrons and holes of IGBT can be expressed accurately. In addition, because it is formulated in the device structure parameters only, any extraction works for fitting parameters are not necessary. In Chapter 4, the established compact model was verified for 600V, 1200V, and 3.3kV rated IGBT’s. The carrier distribution profiles and I-V characteristics calculated by the compact model and TCAD device simulation were compared. As a result of the verification, consistent with high precision characteristics and carrier density distribution, the validity of the model was shown. TCAD simulation needs time-consuming calculation of more than a million times compared to the compact model because it solves all of the basic equations of the semiconductor. In Chapter 5, a scaling principle for the IGBT was established for the first time. The principle was founded from the analysis of the Quasi-2D MOS-ADE model. A roadmap for achieving the next generation power devices was revealed by the principle. Conventional IGBT uses 0.5 to 1 micron design rules, and miniaturization was not considered. The scaling principle opened the way to the adoption of large sized wafers and high resolution semiconductor processes, and showed an explicit direction to the realization of next-generation power devices that improve the production and performance. In Chapter 6, the future of IGBT was discussed based on the knowledge gained in this study.