Si, SiC, and GaN for Power Devices, Part One: Electron Energy and the Semiconductors

The first part of our series contrasting power devices of silicon, silicon carbide, and gallium nitride.

Which semiconductor is the power powerhouse? In this four-part series, we’ll take an in-depth look at the differences between silicon (Si), silicon carbide (SiC), and gallium nitride (GaN) to understand which is best for power devices and why.

In part one, we’ll present a brief review of the nature of electron energy and introduce our semiconductor contenders. We’ll define the band gap, which is used to classify materials as conductors, semiconductors, and insulators. Semiconductors like SiC and GaN are classified as having a wide band gap, which gives them extremely useful properties we will explore later on.

Introduction to Power Electronics

Power electronics applications abound. Application areas like renewable energy (e.g., string inverters, microinverters, and DC optimizers), all-electric aircraft (including eVTOL, electric vertical takeoff and landing aircraft), electric vehicles, and the attendant electric vehicle charging stations demand that the capabilities of power devices evolve. Power devices invariably need to become faster, more efficient (and therefore run cooler), smaller, and ultimately become less expensive.

Silicon carbide (SiC) MOSFETs (Metal Oxide Semiconductor Field Effect Transistors) and gallium nitride (GaN) FETs address these requirements and are replacing the legacy silicon Super Junction MOSFETs and IGBTs (Insulated Gate Bipolar Transistors). Among the advantages provided by SiC and GaN MOSFETS are a wide band gap and high electron mobility. You are likely to encounter the acronyms WBG (Wide Band Gap) and HEMT (High Electron Mobility Transistors) in technical literature. A brief review of electron energy is included in the next section.

A Brief Review of Electron Energy and Bandgap

Figure 1 reminds us that the energy possessed by an orbiting electron increases as it moves away from the nucleus. In this instance an orbital radius can be equated to the energy level.

Figure 1. (Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists).

Figure 1. (Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists).
Figure 2. (Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists).

Figure 2. (Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists).

Figure 2 indicates that orbital electrons must absorb energy to move to a higher energy orbit. Conversely, the orbital electrons must release energy to move to a lower orbital radius. Isolated atoms have discrete energy levels. The formation of energy bands is illustrated in Figure 3.

Energy bands arise when atoms are close together. When two atoms are brought close to each other, an intermixing of electrons in the valence shell will occur. As a result, several permissible energy levels are formed, which is called an energy band.

Figure 3. (Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists).

Figure 3. (Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists).

In Figure 4 we see the development of the energy diagram. An electron volt (eV) is a fundamental unit of energy. It is the energy required to move single electron through a potential difference of one volt and is equivalent to 1.602 x 10-19 joules (J). The valence energy band contains the outermost orbital electrons that are attached to the parent atom. If an orbital electron gains enough energy, it can break free of its parent atom. The electron is then described as being in the conduction energy band. These electrons are classified as being “free”. Free electrons support current flow.

Figure 4. (Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists).

Figure 4. (Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists).
Figure 5. (Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists).

Figure 5. (Used with author’s permission from Discrete and Integrated Electronics Analysis and Design for Engineers and Engineering Technologists).

The band gap is the energy difference between the valence and conduction energy bands. In the discussion that follows, we shall contrast silicon (Si) with silicon carbide (SiC) and gallium nitride (GaN). Figure 5 indicates SiC and GaN possess much larger bandgaps than Si.

Basic Contrast Between Si, SiC and GaN

In addition to bandgap, other important factors to consider are the (maximum) electric field intensity (also called the critical breakdown voltage), the thermal conductivity, the melting point, and the electron (saturation) velocity.

The maximum electric field intensity or the critical breakdown voltage establishes the maximum voltage that can be impressed across a semiconductor before avalanche breakdown occurs. Avalanche breakdown occurs when the electric field intensity is strong enough to rip valence electrons away from their parent atoms.

Thermal conductivity is the rate at which heat passes through a specified semiconductor. It is usually expressed as the amount of heat that flows per unit time through a unit area with a temperature gradient of one degree per unit distance. A semiconductor with high thermal conductivity rids heat quickly. This means they can operate at higher power levels. Alternatively, a device can be made smaller, which reduces its form factor.

Saturation velocity is the maximum velocity a charge carrier in a semiconductor, generally an electron, attains in the presence of high electric fields. A large saturation velocity means a given device can operate at higher frequencies and switch faster.

Carrier mobility is a measure of how quickly a charge carrier (hole or electron) can move through a semiconductor when being pulled (or pushed) by an electric field.

Table 1 provides these values for Si, SiC and GaN.

Table 1. Comparison Between Si, SiC, and GaN.

Table 1. Comparison Between Si, SiC, and GaN.

These values can be plotted using a radar chart as shown in Figure 6 (radar charts are an effective way to present multivariable data). In this instance SiC seems to be the superior semiconductor in the area of thermal conductivity. GaN leads in all other areas. Legacy silicon is largely outmatched.

 Figure 6. Using a Radar Chart to Compare the Si, SiC, and GaN Semiconductors.

Figure 6. Using a Radar Chart to Compare the Si, SiC, and GaN Semiconductors.

Review and Conclusions

Having reviewed electron energy, we are reminded the basic unit of energy is the electron volt (eV). Band gap is the energy difference between the valence and conduction energy bands. Insulators have a band gap that is greater than 5 eV. Semiconductors have a band gap that is less than 5 eV. In conductors, the valence and conduction energy bands overlap, which gives them a bandgap of 0 eV. Conventional semiconductors have bandgaps that range from 0.5 to 1.5 eV. Wide bandgap semiconductors have bandgaps that range from 2 to 4 eV.

Reflection on the comparisons between SiC and GaN leads us to some initial observations.

  • The carrier mobility, bandgap, and electron velocity of GaN exceed the values inherent in SiC. Therefore, GaN has a higher speed capability than SiC.
  • The bandgap and maximum electric field intensity of GaN surpasses the values associated with SiC. GaN can operate at higher voltages than SiC.
  • SiC has a higher thermal conductivity and melting point than GaN. This means SiC is better at operating at higher temperatures than GaN.
  • The speed and voltage capabilities of SiC are less, but close to, GaN.

In part two, we’ll examine the Si power MOSFET, the Si IGBT, the Si super junction MOSFET, and the SiC power MOSFET.