What is Gallium Nitride, GaN Semiconductor

Gallium Nitride, GaN technology is being used increasingly within many areas of electronics design, so understanding what the material actually is can be very useful.

Semiconductors Includes:
What is a semiconductor Holes & electrons Semiconductor materials Compound semiconductors Silicon carbide, SiC Gallium nitride, GaN

Gallium nitride, GaN technology is being used increasingly as a semiconductor material as it is possible to fabricate a number of devices with performance that cannot be equalled by silicon.

In particular GaN technology is used in light emitting diodes where it gives off a blue light and it has been the cornerstone of Blu-ray discs and has given its name to the technology.

But gallium nitride is also being increasingly used in semiconductor power devices, RF and microwave semiconductor components as well as lasers and photonics.

Devices including GaN FETs or GaN HEMTs are increasingly being seen for both RF design and general power electronics applications.

Over the coming years, it is anticipated that GaN technology will increase in its use because of the demand for higher power, higher performance semiconductors will increase. Everything from power circuits to RF designs, LEDs and more.

What is gallium nitride, GaN - physical properties

Gallium nitride has many unique properties as a material. Chemically it comprises gallium which has an atomic number of 31 and nitrogen with an atomic number of 7.

The two elements combine to form what is known as a robust Wurtzite structure. The material is very strong and has a melting point of around 2500°C.

Gallium nitride, GaN does not occur naturally in nature and has to be synthesised chemically. This is achieved by taking gallium and ammonia and subjecting them to high temperatures and pressure.

However there are issues with the material because there are limitations to the size and purity achievable. It has a complex crystal structure known as a Wurtzite structure, and as a result, it is prone to high dislocation densities ranging from 1 in 10⁸ to 1 in10¹⁰ defects.

A summary of the physical properties of gallium nitride are given in the table below.

Main Physical Properties of Gallium Nitride
Properties	Units	Silicon Carbide
Density	g/cm³	6.1
Hardness	GPa	approx 12
Fracture Toughness	MPa.m^1/2	Approx 0.80
Thermal Expansion Coefficient	10^-6/°K
Thermal conductivity	W cm^-1 °K^-1>	1.3
Refractive index	Not applicable - ratio only	2.429
Solubility	Not applicable	Insoluble
Melting point	°C	~2500

What is gallium nitride - electrical properties

Gallium nitride also has many electrical properties as a semiconductor that make it very useful and as a result it is used in many devices from GaN Schottky diodes, GaN FETs, GaN MOSFETs, GaN HEMTS and the like.

Gallium nitride is referred to as a wide bandgap semiconductor material. It has a hexagonal crystal structure. In terms of its semiconductor properties the bandgap is the energy needed to free an electron from its orbit around the nucleus. With a 3.4 eV, the bandgap, gallium nitride is over three times that of silicon, this means that it is often termed a wide bandgap semiconductor or it may just be referred to as "WBG."

Main Electrical Properties of Gallium Nitride Used in Semiconductor Devices
Property	Si	GaN
Energy Band Gap: E_G(eV)	1.12	3.2
Electron Mobility: µ_n(cm²/V_S)	1400	1250
Breakdown Field: E_B(V/cm)X10⁶	0.3	3
Saturation Drift Velocity: v_s(cm/s)X10⁷	1	2.7
Relative Dielectric Constant: e_S	11.8	9.5

Gallium nitride is a group III / V direct bandgap semiconductor. Its high temperature capability as well as its high electron mobility mean that it is ideal for use in many semiconductor devices including GaN FETs, GaN MOSFETs GaN HEMTs and many more devices.

Looking at this further, all semiconductor materials have what is termed a bandgap. The bandgap is an energy range in a solid where no electrons can exist. In other words a bandgap is related to how well a solid material can conduct electricity. Silicon has a bandgap of 1.12 eV whereas gallium nitride has a 3.4 eV bandgap, and silicon carbide has a bandgap of 3.26eV*.

Semiconductors that have a large or wide bandgap are known as wide bandgap semiconductors. The wider bandgap means the semiconductor can sustain higher voltages and higher temperatures. This wide bandgap enables gallium nitride and other wide bandgap semiconductors to be used in many power applications: power supplies as well as RF power amplifiers because they offer high levels of efficiency as well as the greater levels of resilience.

Doping gallium nitride

Obviously one of the key issues associated with GaN technology is the way in which the material is doped. The semiconductor needs to be doped with a suitable material so that it becomes conductive otherwise it will remain an insulator.

Both P-type and N-type materials are needed, although where only one type of material is required, N-type is used because it uses electrons rather than holes and electron mobility is much higher than that of holes.

Normally only a limited number of elements that the GaN crystalline structure will accept to give the P-type or N-type material.

N-type GaN: Silicon or germanium can be used as the dopants to produce N-type gallium nitride.
P-type GaN: The process for producing P-type gallium nitride is more difficult to manage. Magnesium is the only impurity that can produce P-type GaN. The process for this suffers from significant limitations. Magnesium has a much larger ionisation energy than typical dopants used in conventional semiconductor technology. This means that high concentrations of magnesium impurities are required to achieve hole concentrations that are needed for usable devices. In addition to this, high-temperature post-growth annealing is necessary to activate magnesium acceptors, which are passivated by hydrogen in typical growth processes. Overall this process is difficult to satisfactorily achieve.

Although GaN technology is well advanced in many areas, the difficulties with producing the P-type semiconductor has limited some areas of the uptake of GaN technology.

GaN manufacturing

In recent years the manufacture of gallium nitride, GaN technology has moved forward tremendously as the demand for these devices has increased.

Some early GaN FETs were produced around 2006 and these were enhancement mode devices.

The main issues with GaN semiconductor technology are the growth of the raw GaN crystals. For silicon a process known as the Czochralski process is used to draw the silicon crystals, but this is not possible using GaN.

Instead chemical vapour deposition needs to be used to grown the gallium nitride onto a pseudo substrate. This adds time and cost to the process and it does not produce as much gallium arsenide as using the Czochralski process.

Once the substrate has been formed it is possible to produce GaN FETs, GaN FETs, etc using some of the same processes that are used for silicon devices. This means that the cost differential between silicon and GaN is not as large as it might otherwise be.

Toady a good variety of GaN semiconductor devices are produced from GaN FETs (also referred to as GaN HEMTs or just GaN transistors) through to GaN Schottky diodes and also a number of innovative photonic devices as well.

GaN manufacturing processes often use either silicon or silicon carbide as the substrate, i.e. GaN on Si or GaN on SiC.

Each substrate has its own advantages and disadvantages. Silicon is the less costly option compared to silicon carbide, but GaN on SiC offers higher reliability and power. Accordingly GaN on SiC is the approach of choice for many RF power and other applications.

GaN technology advantages & disadvantages

As with anything, there are advantages and disadvantages to using GaN technology. Fortunately most points are advantages, so many circuits are now using devices such as GaN MOSFETs and other GaN technology.

GaN technology advantages

Wide bandgap: Wide bandgap semiconductors permit devices to operate at much higher voltages, frequencies, and temperatures than conventional semiconductor materials.
High breakdown: The wide bandgap of the GaN semiconductors enables them to have a high breakdown voltage. This allows them to be used in high voltage and high power systems more easily.
High electron mobility: The high electron mobility enables the device to respond more quickly giving it a high frequency of operation, faster switching speed and faster transition between on and off states.
Increased switching speed: The higher switching speed and hence the higher frequency of operation for switching within SMPS circuits, etc, results in the use of smaller inductors and capacitors in power circuits. The inductance and capacitance scale down in proportion to the frequency. This can result in a very large decrease in weight and volume, and in turn this reflects in the cost.
Lower system costs: GaN semiconductors are normally higher cost than their silicon equivalents, but it is often possible to gain system level cost reductions by using GaN devices because it is often possible to reduce the size of units, and the increased efficiency and speed (including the frequency of operation) often means that passive components including inductors and capacitors can be smaller and hence cheaper. Often the overall savings can be between 10 and 20%.

GaN technology disadvantages

Issues with P-type material: One of the issues with GaN semiconductors is that of producing the P-type material. Although this does not affect all areas of GaN technology, it does present some major issues in some areas. It means that GaN is not viable for logic where complementary outputs are needed for speed, bipolar transistors where both N-type and P-type semiconductor is required and other devices where P-type semiconductor is needed.
Substrate: Gallium arsenide is very stable and this makes it exceedingly difficult to grow crystals of the material in the same way that it is for silicon. Instead GaN technology uses techniques where the gallium arsenide is grown on what are termed pseudo-substrates. This technique introduces further stages and difficulty into the production of GaN devices.
Substrate dislocations and defects: Another issue with the fabrication of GaN devices is the number of defects that occur in the crystal. These occur and occur at orders of magnitude higher than those with silicon. These give rise to recombination and leakage paths within the GaN devices.
Thermal conductivity: Although gallium arsenide has a similar thermal conductivity to that of silicon, the fabrication of GaN devices such as GaN FETs and Gan diodes, etc, these involve the inclusion of layers of layers of materials such as alloys of GaN and aluminium nitride AlN. These have very poor levels of thermal conductivity and this can be a limiting factor in many aspects of the use of GaN technology rather than its electrical performance.
Fabrication costs: In view of the difficulties in the production of devices using GaN technology including GaN diodes and GaN MOSFETs, etc, the costs of the devices tend to be much higher. As a result, circuit designers need to weigh up the cost / performance benefits of using gallium arsenide devices. That said in many instances the use of devices using GaN technology can result in an overall system cost reduction, but care needs to be taken to ensure these benefits are fulfilled in the circuit design.

The use of GaN semiconductors in circuits can have a very positive impact on the overall electronic circuit design, although the issues with the fabrication of devices using GaN technology can result in higher costs, although these can often be offset against the overall benefits and bring a better and lower cost finished design.

The use of GaN semiconductors is growing in many areas of electronics. As a result of its advantages over silicon in many applications, it able to provide some significant improvements in performance. As a result the development of GaN technology is moving forward and many hurdles are likely to be overcome leading to a more widespread use of the technology.