Why the GaN bandgap is important

The GaN bandgap is one of the key reasons why the GaN material has become a leading choice for modern power electronics. With a bandgap of 3.4 eV, GaN can withstand higher voltages, switch faster, and operate at higher temperatures than conventional silicon devices. Understanding the GaN bandgap helps explain the performance advantages that make GaN a key technology in modern power conversion systems.

Semiconductor band structure: valence band, conduction band and bandgap

Valence band vs. conduction band

In solid-state physics, the electrons inside a material occupy specific energy levels that are grouped into two main bands. The valence band is the lower-energy band. It is completely filled with electrons that are tightly bound to their parent atoms, so these electrons cannot move freely and cannot conduct electricity.

The conduction band is the higher-energy band, and it is mostly empty. Once an electron manages to jump up into it, that electron can move freely through the material. This is what creates an electric current.

Between these two lies the bandgap, also called the forbidden gap: a range of "forbidden" energy levels that electrons cannot occupy. An electron must either stay below the gap in the valence band or gain enough energy to jump completely across it into the conduction band.

Diagram of a semiconductor band structure showing electrons in the conduction band, holes in the valence band, and the bandgap between the two energy bands.

Diagram of a semiconductor band structure showing electrons in the conduction band, holes in the valence band, and the bandgap between the two energy bands.
Diagram of a semiconductor band structure showing electrons in the conduction band, holes in the valence band, and the bandgap between the two energy bands.

Diagram of a semiconductor band structure showing electrons in the conduction band, holes in the valence band, and the bandgap between the two energy bands.
Diagram of a semiconductor band structure showing electrons in the conduction band, holes in the valence band, and the bandgap between the two energy bands.

Diagram of a semiconductor band structure showing electrons in the conduction band, holes in the valence band, and the bandgap between the two energy bands.
energy-band-structures-conductors-semiconductors-insulators

Energy band structures of conductors, semiconductors and insulators
energy-band-structures-conductors-semiconductors-insulators

Energy band structures of conductors, semiconductors and insulators
energy-band-structures-conductors-semiconductors-insulators

Energy band structures of conductors, semiconductors and insulators
Diagram comparing the energy band structures of a conductor, semiconductor, and insulator, illustrating the increasing bandgap between the valence and conduction bands.

Diagram comparing the energy band structures of a conductor, semiconductor, and insulator, illustrating the increasing bandgap between the valence and conduction bands.
Diagram comparing the energy band structures of a conductor, semiconductor, and insulator, illustrating the increasing bandgap between the valence and conduction bands.

Diagram comparing the energy band structures of a conductor, semiconductor, and insulator, illustrating the increasing bandgap between the valence and conduction bands.
Diagram comparing the energy band structures of a conductor, semiconductor, and insulator, illustrating the increasing bandgap between the valence and conduction bands.

Diagram comparing the energy band structures of a conductor, semiconductor, and insulator, illustrating the increasing bandgap between the valence and conduction bands.

Classifying materials by bandgap size: conductors, semiconductors and insulators

The size of the bandgap — measured in electronvolts (eV) — decides whether a material behaves as a conductor, a semiconductor or an insulator.

In conductors such as copper and aluminium, the valence and conduction bands overlap. There is effectively no bandgap (0 eV), so electrons move freely at any temperature. This is what makes these materials excellent conductors. In insulators such as glass and quartz, the bandgap is exceptionally large (above roughly 5 eV). Forcing an electron across it would take an unmanageable amount of energy or extreme heat.

As a result, insulators do not conduct under normal conditions. Semiconductors such as silicon and germanium sit between the two, with a moderate bandgap (typically around 1 to 1.5 eV). At absolute zero they behave like insulators. At room temperature, though, a small amount of heat, light or applied voltage is enough to push electrons across the gap. This lets the material conduct on demand.

The gallium nitride bandgap: a 3.4 eV direct wide bandgap

Controlling and exploiting the bandgap is the foundation of all modern electronics. Semiconductors bridge the gap between conductors and insulators. Engineers can therefore use external forces, such as a transistor gate voltage, to control whether current flows through the device. The width of the gap sets the limits of that control. Standard silicon has a bandgap of about 1.1 eV. If a silicon chip gets too hot — typically above 150 °C — ambient thermal energy accidentally kicks too many electrons into the conduction band. The chip then loses its ability to turn off, leading to uncontrollable leakage current and system failure.

This is why gallium nitride (GaN) — a compound semiconductor made from gallium and nitrogen — is so revolutionary. With a bandgap of 3.4 eV, its bandgap barrier is roughly three times larger than silicon's. As a result, GaN needs much higher voltages or temperatures before it fails accidentally. It can therefore handle massive power loads, switch at extreme speeds and run hot without melting down.

Device implications of the wide GaN bandgap

The size of the bandgap is directly tied to a material's critical electric-field breakdown. When a high voltage is applied across a semiconductor, electrons accelerate. If they move fast enough, they slam into atoms and knock more electrons free. This creates an uncontrollable chain reaction known as avalanche breakdown. In a wide-bandgap material like GaN (3.4 eV), electrons must gain significantly more energy before they can bridge the gap and trigger this effect.

As a result, GaN can withstand an electric field about 10 times stronger than silicon before failing. This lets engineers design high-voltage chips that are incredibly thin, minimising internal resistance (RDS(on)) and saving large amounts of power.

As a semiconductor heats up, ambient thermal energy naturally lifts electrons from the valence band into the conduction band, creating leakage current. Because silicon's bandgap is relatively narrow (1.1 eV), it takes very little heat to do this: at around 150 °C to 175 °C the leakage becomes so severe that the transistor can no longer turn off, the chip suffers thermal runaway and fails. With its 3.4 eV barrier, GaN does not let ambient heat force electrons across the gap so easily.

GaN devices can therefore operate reliably at temperatures well above 200 °C to 300 °C. In industrial power grids, electric vehicles and aerospace systems, this eliminates the need for heavy, expensive liquid-cooling systems or massive aluminium heatsinks.

The bandgap also dictates how a transistor behaves when it transitions between its on and off states. Because wide-bandgap materials can block high voltages using much thinner internal layers, the transistor's physical dimensions can be shrunk significantly. Those smaller structures mean drastically lower internal parasitic capacitances (CISS, COSS, CRSS).

With less capacitance to charge and discharge on every cycle, GaN transistors can switch on and off up to 10 to 100 times faster than silicon. Faster switching allows smaller passive components such as inductors and capacitors. That is why a modern GaN laptop charger can deliver 100 W while still fitting easily into a pocket.

The size of the bandgap also determines exactly which wavelength — and therefore colour — of light a semiconductor will interact with. When an electron drops from the conduction band back down to the valence band, it releases its excess energy as a photon. The energy of that photon perfectly matches the bandgap energy (E = hν).

Silicon's bandgap properties cause it to release this energy mostly as heat rather than light, because it has an indirect bandgap. GaN, by contrast, has a direct bandgap of 3.4 eV, which corresponds precisely to high-energy blue and ultraviolet light.

The discovery of the GaN-based blue LED was recognised with the 2014 Nobel Prize in Physics. It is what made modern energy-efficient white LED lighting, Blu-ray technology and vibrant smartphone displays possible.

The size of the bandgap is directly tied to a material's critical electric-field breakdown. When a high voltage is applied across a semiconductor, electrons accelerate. If they move fast enough, they slam into atoms and knock more electrons free. This creates an uncontrollable chain reaction known as avalanche breakdown. In a wide-bandgap material like GaN (3.4 eV), electrons must gain significantly more energy before they can bridge the gap and trigger this effect.

As a result, GaN can withstand an electric field about 10 times stronger than silicon before failing. This lets engineers design high-voltage chips that are incredibly thin, minimising internal resistance (RDS(on)) and saving large amounts of power.

As a semiconductor heats up, ambient thermal energy naturally lifts electrons from the valence band into the conduction band, creating leakage current. Because silicon's bandgap is relatively narrow (1.1 eV), it takes very little heat to do this: at around 150 °C to 175 °C the leakage becomes so severe that the transistor can no longer turn off, the chip suffers thermal runaway and fails. With its 3.4 eV barrier, GaN does not let ambient heat force electrons across the gap so easily.

GaN devices can therefore operate reliably at temperatures well above 200 °C to 300 °C. In industrial power grids, electric vehicles and aerospace systems, this eliminates the need for heavy, expensive liquid-cooling systems or massive aluminium heatsinks.

The bandgap also dictates how a transistor behaves when it transitions between its on and off states. Because wide-bandgap materials can block high voltages using much thinner internal layers, the transistor's physical dimensions can be shrunk significantly. Those smaller structures mean drastically lower internal parasitic capacitances (CISS, COSS, CRSS).

With less capacitance to charge and discharge on every cycle, GaN transistors can switch on and off up to 10 to 100 times faster than silicon. Faster switching allows smaller passive components such as inductors and capacitors. That is why a modern GaN laptop charger can deliver 100 W while still fitting easily into a pocket.

The size of the bandgap also determines exactly which wavelength — and therefore colour — of light a semiconductor will interact with. When an electron drops from the conduction band back down to the valence band, it releases its excess energy as a photon. The energy of that photon perfectly matches the bandgap energy (E = hν).

Silicon's bandgap properties cause it to release this energy mostly as heat rather than light, because it has an indirect bandgap. GaN, by contrast, has a direct bandgap of 3.4 eV, which corresponds precisely to high-energy blue and ultraviolet light.

The discovery of the GaN-based blue LED was recognised with the 2014 Nobel Prize in Physics. It is what made modern energy-efficient white LED lighting, Blu-ray technology and vibrant smartphone displays possible.

Gallium nitride vs. silicon: device physics and figures of merit

The bandgap dictates how silicon and GaN devices are physically built to block voltage. To block high voltages such as 650 V, a silicon MOSFET needs a thick internal vertical "drift region." Because silicon's critical breakdown field is low (0.3 MV/cm), this layer has to be wide to prevent electrical arcing. A thicker layer forces electrons along a longer, higher-resistance path, which balloons the device's on-resistance (RDS(on)).

However, a GaN eHEMT works differently. GaN's wide bandgap yields a much higher critical breakdown field (3.3 MV/cm). It can therefore block that same 650 V across a layer about 10 times thinner, which allows the device to be built laterally. The wide bandgap enables a 2DEG (two-dimensional electron gas): a highly concentrated, almost frictionless sheet of electrons at the AlGaN/GaN interface, through which electrons move horizontally with virtually zero obstruction.

For power electronics engineers, the primary measure of transistor quality is the Figure of Merit (FoM). Common examples are RDS(on) x Qg or RDS(on) x QOSS, where Qg is the total gate charge and QOSS the output charge. In silicon, achieving a low RDS(on) means making the chip physically larger, but a larger chip automatically raises the internal capacitances (CISS, Coss).

That demands more charge (Qg) to switch the device, which slows switching and spikes energy losses. In a GaN eHEMT, the wide bandgap lets a tiny, ultra-thin lateral channel reach an extremely low RDS(on).

As a result, the chip can be miniaturised and its parasitic capacitances fall dramatically. GaN can achieve a Figure of Merit up to 10 to 100 times better than silicon. This enables switching frequencies in the tens of megahertz (MHz) range while keeping losses extremely low.

In half-bridge topologies, common in motor drives and power supplies, current is periodically forced to flow backward through the transistor.

A silicon MOSFET has an inherent parasitic body diode built into its PN-junction structure. When current reverses, this body diode conducts. Then, when the transistor tries to turn off, it must first clear the stored minority carriers from the diode. That delay creates a large spike of wasted energy called reverse recovery charge (Qrr).

A GaN eHEMT has no physical body diode, thanks to its lateral, majority-carrier 2DEG channel. It conducts in reverse through the same 2DEG channel when biased correctly, so its Qrr is exactly zero. This removes the leading cause of switching failure and overheating in high-frequency hard-switched circuits.

The bandgap dictates how silicon and GaN devices are physically built to block voltage. To block high voltages such as 650 V, a silicon MOSFET needs a thick internal vertical "drift region." Because silicon's critical breakdown field is low (0.3 MV/cm), this layer has to be wide to prevent electrical arcing. A thicker layer forces electrons along a longer, higher-resistance path, which balloons the device's on-resistance (RDS(on)).

However, a GaN eHEMT works differently. GaN's wide bandgap yields a much higher critical breakdown field (3.3 MV/cm). It can therefore block that same 650 V across a layer about 10 times thinner, which allows the device to be built laterally. The wide bandgap enables a 2DEG (two-dimensional electron gas): a highly concentrated, almost frictionless sheet of electrons at the AlGaN/GaN interface, through which electrons move horizontally with virtually zero obstruction.

For power electronics engineers, the primary measure of transistor quality is the Figure of Merit (FoM). Common examples are RDS(on) x Qg or RDS(on) x QOSS, where Qg is the total gate charge and QOSS the output charge. In silicon, achieving a low RDS(on) means making the chip physically larger, but a larger chip automatically raises the internal capacitances (CISS, Coss).

That demands more charge (Qg) to switch the device, which slows switching and spikes energy losses. In a GaN eHEMT, the wide bandgap lets a tiny, ultra-thin lateral channel reach an extremely low RDS(on).

As a result, the chip can be miniaturised and its parasitic capacitances fall dramatically. GaN can achieve a Figure of Merit up to 10 to 100 times better than silicon. This enables switching frequencies in the tens of megahertz (MHz) range while keeping losses extremely low.

In half-bridge topologies, common in motor drives and power supplies, current is periodically forced to flow backward through the transistor.

A silicon MOSFET has an inherent parasitic body diode built into its PN-junction structure. When current reverses, this body diode conducts. Then, when the transistor tries to turn off, it must first clear the stored minority carriers from the diode. That delay creates a large spike of wasted energy called reverse recovery charge (Qrr).

A GaN eHEMT has no physical body diode, thanks to its lateral, majority-carrier 2DEG channel. It conducts in reverse through the same 2DEG channel when biased correctly, so its Qrr is exactly zero. This removes the leading cause of switching failure and overheating in high-frequency hard-switched circuits.

Key takeaway on the GaN bandgap

Gallium nitride (GaN) outperforms traditional silicon (Si) in power electronics, driven primarily by its larger bandgap energy of 3.4 eV versus 1.1 eV. This fundamental advantage in physics lets GaN devices handle higher voltages, switch significantly faster and minimise power losses.

See more insights about GaN from a global leader in power semiconductors

Explore more about GaN

Frequently asked questions about GaN Bandgap

A semiconductor bandgap is the energy difference between the top of the valence band and the bottom of the conduction band. To carry an electric current, an electron must gain at least this much energy and cross the gap. The size of the bandgap therefore determines whether a material conducts — and how easily.

The forbidden gap, also known as the bandgap, is the energy range between the valence band and the conduction band where no allowed electron energy states exist. Electrons cannot occupy this region and must gain enough energy to cross it before they can contribute to electrical conduction.

Electrons cross the bandgap by absorbing external energy — typically heat, light (photons), or the energy supplied by an applied electric field. Only when the absorbed energy is at least as large as the bandgap can an electron move up into the conduction band and contribute to conduction.

A material is classed as wide-bandgap when its energy gap exceeds roughly 3.0 electron volts — gallium nitride (GaN) being a classic example.

Bandgap energy is measured in electronvolts (eV). One electronvolt is the energy an electron picks up as it moves across a one-volt difference. Because the energy gaps in semiconductors are so small, this tiny unit is far more practical than the joule.

In conductors, the valence and conduction bands overlap, so there is effectively no bandgap and electrons move freely. Semiconductors have a moderate bandgap, typically between about 0.5 and 3.0 eV, so they conduct only when supplied with extra energy. Insulators have a large bandgap, usually above roughly 4.0 eV, which prevents conduction at normal operating temperatures and voltages.

A wide-bandgap semiconductor is a material whose energy gap is significantly larger than silicon's. The gallium nitride semiconductor (GaN) is a leading example: its wider bandgap enables higher breakdown voltages, lower switching losses, and reliable operation at higher temperatures and frequencies than silicon can sustain.

At room temperature it measures about 3.4 eV, compared with roughly 1.12 eV for silicon. This nearly threefold difference is the fundamental reason GaN devices can block higher voltages, switch faster, and tolerate more heat than their silicon counterparts.

Silicon carbide (4H-SiC) has a bandgap of about 3.26 eV, while gallium nitride (GaN) is about 3.4 eV. Both are wide-bandgap semiconductors — far above silicon's roughly 1.1 eV — which is why both outperform silicon in high-voltage, high-temperature, and high-frequency power electronics.