Bipolar Junction Diode
In Chapter 2, it has given brief description on the p-n junction as a two-terminal semiconductor bipolar device functioning for rectification. The junction acts as a switch which allows current to flow under forward bias condition and blocks the current by a much higher potential barrier under reverse bias condition. So, a simple p-n junction diode is serving as an uncontrollable switch. In a sense, the behavior of the semiconductor diode is preset by the internal p-n junction, rather than by or with any combined influence of external control signal. The diode structure can be complemented by the field-controlled mechanism, called field-controlled diode, or similar variations to create the ability of external control (Baliga, 1981; Thapar and Baliga, 1997). Bear in mind that, the semiconductor diode can also be used for applications other than current switching, e.g. used as a small signal attenuator with no current switching. Or, for some applications, the diode is fabricated in the integrated circuits serving as a capacitor under reverse bias condition. In this case, the diode will not carry current, but serve as a voltage-controlled junction capacitor. Nevertheless, for the main applications of power semiconductor diodes described here, the main properties remain on the switching behaviors and power related issues, such as the current conduction, blocking and reverse recovery.
There are various ways to form a diode structure on silicon and the variations mainly due to the process differences and the need for integration with other devices on the same silicon die. Sometimes, the diode structure is formed as a parasitic component integrated with another device, such as MOSFET device has a built-in body diode. Figure 3.1 shows various diode structures formed by different fabrication processes. Among them, Fig. 3.1(a) shows a simple vertical p-n junction diode by diffusion process. Figure 3.1(b) is the
Fig.3.1. Various diode structures: (a) vertical p-n diode, (b) vertical p-i-n diode, (c) lateral p-n diode, (d) lateral p-i-n diode, and (e) MOSFET body diode.
vertical p-i-n diode using the similar fabrication process as in (a), but made on the p+-n- epi-wafer. Figure 3.1(c) gives the lateral p-n diode where the p+ layer is very shallow made for ohmic contact purpose. The lateral p-n diode is suitable for integration. Figure 3.1(d) gives the lateral p-i-n diode using the n-well in CMOS process as the drift region. Figure 3.1(e) is a lateral LDD n-MOSFET structure with p-body shorted to the source (left n+ contact). The body diode is then formed between the p-body and the drain contact (right n+ contact) with the n-well drift region in between. The body diode may conduct during the fly-wheeling process used in power converters.
The behaviors of the junction diode device are determined by two basic elements, namely the junction and the semiconductor doping within the device. The junction behaves like a check-valve, which allows a certain type of carriers to go through but not the other type of the carriers in the same direction of flow. To describe the concept more clearly, let us take a p+-n junction as an example. If the junction is reverse biased, then the electric field is running from the n-side to the p+-side. Under such a condition, the junction will allow electron carriers to move from p+-side to n-side, or equally capable for hole carriers to move from n-side to p+-side. And at the same time, the electron carriers are prevented to move from n-side to p+-side, nor can the hole carriers flow from the p+-side to n-side. The concept seems to be trivial for one single junction under steady-state, but it becomes rather important when dealing with a multiple-junction device or under transient condition when electron and hole carriers are mixed at different regions and a reverse-biased junction does carry current flow. The second element affecting the behaviors is the doping level. If an area is used for carrier emission, it then needs to have a higher concentration of doping for a better emission efficiency to boost up the current density. Otherwise, for a low emission efficiency it means not enough number of carriers and, in term, the device has a high on-state voltage drop. If the area serves to sustain high field, then a lower doping level drift-region shall be used to raise the breakdown voltage. The high resistivity of a lightly doped region at off-state may not be the same resistivity during on-state current conduction depending on the amount of excess carriers flowing into this region. If a large amount of excess carriers are injected into the lightly doped region, then the injection creates an effect called the conductivity modulation for the situation that more free carriers are now available within the region and the resistivity of the region becomes much lower for the conduction period.
In modern power electronic circuits, the diode device plays a major role in current rectification and energy fly-wheeling. A fast switching diode leads to a lower switching loss and in turn a higher system efficiency. Besides the switching speed, other desired characteristics such as low forward voltage drop, low leakage current, and high temperature capability are also desired. Power Schottky barrier diodes made of metal-silicon junction are generally used for high-speed switching applications due to its less amount of excess carrier storage compared to the conventional diodes. However, the Schottky barrier diode has a limited role above 100 V due to its higher reverse leakage current. The p+-n--n+ (or called p-i-n) diodes can be used for high-voltage switching and rectification purpose. However, the switching speed is relatively slower for the recovery time needed to clear up the excess carriers in the long drift region. High voltage GaAs power diodes are now commercially available for high-speed and high temperature switching applications for its larger bandgap compared with that of silicon. Diodes made of SiC material are still in the laboratory development stage and hopefully to be available soon in the market.
In this chapter, the device physics of power p+-n--n+ and Schottky barrier diodes are reviewed and discussed. An experimental investigation on the high temperature characteristics of a silicon p+-n--n+ diode is accompanied to completely cover the steady-state silicon diode properties. Diodes of GaAs and SiC are briefly described. This is followed by switching behaviors of the junction diodes. New devices, such as the field-controlled diode, MPS (Merged p-(i)-n and Schottky) diode and synchronous rectifier are also described at the end of the chapter.