Ultra-wide Band Gap Semiconductors

[Reddington777]

Starting this thread specifically to follow new technology and updates for ultra-wide band gap semiconductors including but not limited to Ga_2_O3 (gallium oxide), AIN (aluminum nitride), and Diamond.

China has been making strides in this area with reports indicating a demo 8mm wafer size of GaO being developed. I'm not terribly familiar with Chinese advancements in this area so I'll leave it to people who know.

On the US side, AFR has been making strides as well with a number of startups and small companies. A number of contracts for GaO semis have went out through SBIR contracts. AFRL has also been investigating AIN and Diamond based semiconductors. To start off, here is an article from AFRL from 2020 regarding GaO semis:

https://compoundsemiconductor.net/article/111289/AFRL_Breaking_Records_With_Gallium_Oxide/feature

 

[starviking]

What are the specific advantages of ultra-wide band gap semiconductors?

 

[Reddington777]

I'm not really well versed in this and I'm barely barely just reading about it a few months ago but because they have a high band gap, they waste less energy, operate at higher temperatures (less demanding on cooling) and produce greater power while having less leakage.

Wide band gap semis each have their own benefits.

• GaO is good for power switches.
• AIN has the largest bandgap - higher breakdown and temp tolerance and slightly better electron mobility than GaO
• Diamond really is the best of the three in breakdown and thermal conductivity and only slightly behind AIN in bandgap but is exceptionally difficult and expensive to manufacture

Generally, all three are being explored / worked on with GaO being the next most likely to mature. I'd really like to give you a three page long write up about it but semis isn't anywhere close to anything I've studied before so ... unfortunately, it's mainly what I can skim from dumbed down articles - and barely even that.

Certainly hope someone more educated in this area than I would speak more about it.

 

[Nik]

"Diamond really is the best of the three..."
Should they mention mono-isotopic diamond, they're getting seriously serious...

Apparently its thermal conductivity is just that little bit better than the 'natural' mix...

 

[OliverSedlacek]

Not particularly high bandgap but this caught my eye InSe silicon alternative

 

[Reddington777]

https://thedefensepost.com/2026/08/24/bae-darpa-rf-heat-reduction/amp/

DARPA laying the ground work to build cooling into semi conductor devices architecture. They are trying to reach "16x the output power density of production GaN amplifiers".

Though tangentially related, this could possibly be laying the groundwork for future UWB semi devices.

If anyone wants to dive in to the official docs and try to make more sense of them they are here:

https://sam.gov/opp/efe061b616704f59a798c9d72163e678/view?

 

[Scott Kenny]

https://thedefensepost.com/2026/08/24/bae-darpa-rf-heat-reduction/amp/

DARPA laying the ground work to build cooling into semi conductor devices architecture. They are trying to reach "16x the output power density of production GaN amplifiers". Im not able to find the original DARPA

The good trick would be thermal management that isnt electrically conductive or even semiconductive. That would let you lay down thermal management layers directly onto the chips electrical connections.

For that matter, if someone could finally figure out thermal superconductors, that would be nice. (I'm fairly reliably informed that they aren't prohibited by the laws of physics)

 

[OliverSedlacek]

Thermal conductor electrical insulator, can I interest you in my diamond range?

 

[MigLui]

For that matter, if someone could finally figure out thermal superconductors, that would be nice. (I'm fairly reliably informed that they aren't prohibited by the laws of physics)

Superconduction is a result of the change of state ( pairing up for electrical SCs ) of the involved quantum particles, such that they act more like bosons ( obey Bose-Einstein statistics ), and no longer like fermions ( obey Fermi-Dirac statistics ).

What this essentially means, is that the superconducting pairs can inhabit the same state and not interact with each other ( like photons of light that don't interact with each other )

Since temperature transfer is essentially a transfer of kinetic energy between particles, the absence of interaction would not make temperature transfer possible.
That is the basic definition of superconduction; energy is not lost by 'carriers' as they travel through the medium.

I would like to see this 'reliable information' if you could post a link. Thanks.

 

[Nik]

Tangential, I remember SciFi writer Larry Niven being 'taken to task' over his Ringworld tales.

"... the Ringworld is unstable" was the herd of elephants in the pool.

But, he was *also* chastised for trailing a superconductor strand into near-by pool or stream as a beam-weapon shield coolant...

 

[Scott Kenny]

That is the basic definition of superconduction; energy is not lost by 'carriers' as they travel through the medium.

Basic definition of electrical superconduction.

Thermal superconduction uses different mechanisms. And things like highly pure sapphire crystals or carbon nanotubes can move heat nearly lossless as the phonons travel without significant scattering, called ballistic phonon conduction. Though this apparently only applies at annoyingly low temperatures, just like electrical superconductors. Superfluid helium has a different mechanism, called second sound or temperature waves.

Neither one of them makes sense to me. But it works and has physics papers written on it.

 

[Seragina]

Still, carbon is a good candidate as it combines both properties best.
In any case we are talking about RF here which means anything that's prone to recieve RF energy could interfere with the circuits. You may recall carbon/graphic used to serve as antenna in the early days. A significant amount of work goes into noise filter and amplification circuitry for a reason.

 

[MigLui]

Neither one of them makes sense to me. But it works and has physics papers written on it.

Thanks for that; will look into ballistic phonon conduction.

My understanding is that temperature is an 'emergent' property of the statistical motion ( kinetic energy ) of particles.
Temperature is 'transferred' by one particle imparting some kinetic energy to another. This is done through field interactions, until temperature equilibrium, and maximal entropy of the system, is reached.
IOW, one particle cannot impart some energy to another particle without losing, at least that amount, itself ( energy and momentum conservation ). A good heat conductor implies good interaction, unlike electrical superconduction.

Edit: Ballistic phonon conduction - "a heat transport regime where phonons (quantized lattice vibrations) travel across a material without any internal scattering or collisions, moving in straight lines from a hot source to a cold sink"

Yes, heat transfer confined to field interactions.
So, although an efficient conduction mechanism, it is in no way a superconduction mechanism, as it improves energy loss and stops even quicker.

 

[Michel Van]

Ultra-wide band gap semiconductors offer:

High Breakdown Field: Can sustain much higher voltages without electrical breakdown compared to silicon.

Extreme Environment Tolerance: Operate reliably at high temperature1s,
high-frequency radio/microwave bands, and intense radiation fields.

Deep-Ultraviolet (DUV) Transparency: Ideal for short-wavelength optical emission, including deep-UV LEDs and lasers.

Primary Applications would be:

High-Power Electronics: Next-generation grid control, high-voltage switches, and solid-state transformers.

High-Frequency RF Devices: Advanced communication systems and military radar components operating
well into millimeter-wave and higher frequencies.

Optoelectronics: Deep-UV photonic and laser devices.

Sensors & Quantum Information: Harsh-environment sensors and quantum sensing architectures.

one application would Jupiter Probe that operate without problem in Harsh radiation belt.
what use millimeter radar to scan Jupiter large moon and UV laser for high data communication.

 

[Seragina]

Edit: Ballistic phonon conduction - "a heat transport regime where phonons (quantized lattice vibrations) travel across a material without any internal scattering or collisions, moving in straight lines from a hot source to a cold sink"

Yes, heat transfer confined to field interactions.
So, although an efficient conduction mechanism, it is in no way a superconduction mechanism, as it improves energy loss and stops even quicker.

Well, one is through a quantum force field(phonon) and the other is through an actual electron particle (even if the charge can move indepently).

 

[MigLui]

Well, one is through a quantum force field(phonon) and the other is through an actual electron particle (even if the charge can move indepently).

Electrons don't actually have a 'size'.
They are point quantum particles, or manifestations of the QED electron field.

( BSc in Physics )