Semiconductor components with current ratings exceeding 1 ampere are generally classified under the term power semiconductors. The blocking capability of such devices normally ranges from a few volts up to the 10 000 V mark. Most power devices lead an inconspicuous existence. Nevertheless, with the advancement of power electronic systems, it is a fact that they continue to play an ever more important role in technology and in everyone’s life. Power semiconductor components supply our electrical equipment with energy: devices for low power (watts to kW) are found in the power supplies of virtually every electric and electronic device, such as computers, TVs, washing machines, and refrigerators; in the medium power range (10 kW to several MW), they power the motors of locomotives and industrial drives, or they supply smelting furnaces with energy. Ultimately, in the Gigawatt range, very high voltage devices are found in HVDC systems (High-Voltage DC transmission).
Fundamentally, power semiconductors and integrated circuit chips (ICs) are very similar in that they both comprise of pn-junctions, bipolar transistors, and MOS structures. Thus, the basic semiconductor device physics and theory applies to both types of components. However, especially for power devices rated above 600 V, an additional dimension must be considered: in order to improve their characteristics, high voltage devices are flooded under certain operating conditions with plasma, which represents a dense mixture of positively and negatively charged carriers. Such plasmas play an important role in defining the reaction of a semiconductor device to a change in the external conditions. This makes it extremely difficult to fully describe the characteristics and the behavior of a component in a data sheet.
The aim of this book is to provide an overview of the various types of power semiconductor devices, to give an insight into how they function, and to explain and analyze their characteristics. All the important classes of power semiconductors are covered, and thus, one can assume that nearly all components available on the market belong to one of these groups.
Power Semiconductor Materials
For some years, engineers have realized that materials such as Silicon Carbide (SiC), Gallium Nitride (GaN), and Diamond (C) are, in principle, more suitable for power semiconductors than silicon. This is mainly due to the fact that such materials exhibit much higher breakdown field strengths in relation to silicon. Thus, in order to achieve a desired minimum voltage capability, a device thickness significantly below that of silicon is sufficient. Since the thickness of the semiconductor component is the most important factor determining the overall losses, the use of SiC, GaN, and C would enable a quantum leap in high-voltage power electronics evolution.
In particular in the field of SiC, there are currently numerous research projects underway. However, there are still difficult problems to be solved, in particular with regard to material quality. These problems result in a high manufacturing cost of high-quality SiC wafers. Before a large-scale industrial use can be contemplated, the price of the starting material must drop more than an order of magnitude. Silicon, on the other hand, is an extremely well-researched material, and has been produced for some time with an almost perfect quality and at a low cost. Hence, a wide-scale replacement of silicon can be ruled out in the next decade. It is for this reason that this book mainly concentrates on silicon devices. Unless explicitly stated, all the numerical examples are calculated using the properties of silicon.
Fundamentals of semiconductors physics
Foundations for the modern understanding of the structure of matter were laid around the beginning of the 20th century. In 1897, J.J. Thomson discovered the electron. He concluded that the atom consists of negative and positive charge, and suggested the “plum pudding” model of the atom, in which the negatively charged electrons are embedded in a cloud of positive charge. Less than a quarter of a century later, in 1911, Ernest Rutherford discovered in an alpha particle scattering experiment that the positive charge of the atom is concentrated in a very small nucleus, whose size he found to be 10 000 times smaller than that of the atom. He correctly concluded that the electrons orbit the nucleus. However, he believed that they can assume any arbitrary energy. In 1913, Niels Bohr found that this was not correct, and suggested that the electrons in an atom can only take on discrete, well defined energy levels, which he called shells. Bohr’s shell model enabled the understanding of the chemical, physical, and electrical properties of matter.
We will use Bohr’s shell model to explain the basic structure and properties of a silicon crystal. Then, we will gradually introduce all important phenomena that determine the behavior and characteristics of semiconductor devices.
The first section, which discusses the atomic model, requires a brief excursion into quantum mechanics. This might prove challenging for some readers, partly because a detailed explanation of the theory is beyond the scope of this book. However, all subsequent sections of this chapter can be understood without the in-depth knowledge of the atomic model.