Does a wide bandgap JFET glow under forward bias?

Maybe. A SiC item is not just 'wide bandgap', but DIRECT bandgap in nature; there's no highspeed way for multistep recombination to far outpace the one-step-light-emitting that causes the blue glow.

I recall pictures of bipolar transistors emitting weak white light. Pease?

CMOS does it too--google mcmanus "picosecond imaging circuit analysis"

Cheers

Phil Hobbs

torsdag den 17. september 2020 kl. 19.35.47 UTC+2 skrev John Larkin:

with enough power most components do ;)

Pictures, but perhaps taken with an IR camera? There's a good quantum-mechanics reason to expect that a silicon charge-carrier-injection across the bandgap does not make white light, but (of course) an IR camera that picks up light is capable of visible-light in white or false color as a readout option.

Pictures of emssion is possible. White emission is not.

Of course it is.

Reverse biased pn junctions represent a different situation. When a small reverse bias is applied to a junction, the depletion region widens, causing low current but a substantial electric field. As the reverse bias is increased, the probability that a highly excited electron will cross the junction increases. This will generate photons from the recombination of carriers whose energies can be significantly above the bandgap energy. The resulting emission spectrum will have a significant tail which can extend into the visible wavelengths.

The light was emitted by an EB junction in reverse breakdown. Many years ago, after reading Pease's piece on the subject, I opened up a transistor in a TO-18 can to take a look. The light was whitish. That surprised me. I expected monochromatic light.

Jeroen Belleman

With tens or hundreds of volts c-e, I'd expect a normally biased transistor to emit some visible light.

Zeners, too.

Nuts do it as well, so why shouldn't a JFET?

Both effects could probably be combined in a fancy way.

Best regards, Piotr

Yes, that's certainly possible. The top post, though, specified 'forward bias' and adding in a fluorescence due to higher energy electrons (an electron-beam probe could do it, too) is entirely a different kettle of fish.

I have never been able to go past the yellow part of the spectrum, although the achieved quantum efficiency was rather impressive and would not be adequately described by the word "weak"... ;-)

Best regards, Piotr

Look up the PICA paper I posted. Hot-carrier emission is much bluer than that.

Cheers

Phil Hobbs

Ohhhh, you didn't mean peanuts or gonads, it seems.

Back in 2015 in this group I posted these pics:

The e-b in breakdown junction light looked silvery-blue-gray to the human eye but the phone camera image represents a different color so possibly a strong IR component too?

piglet

All sorts of semis may glow, but most are potted in black epoxy.

It would be cool to zener a photodiode. They are specifically designed for good optical paths into (thus out of) the semiconductor.

A forward-biased compound-semi photodiode probably has some, maybe visible, LED effect.

Measuring the photo current, and then displaying the value in mV?

The forward voltage of the photo diode will be very depending on temperature (threshold, leakage will vary). The diode is placed on top of the FET gel, so it's temperature will also have varied.

Apart from the fun fact that light has been observed, this research is practically useless.

It should be mandatory to add a well-qualified electronics engineer to scientific teams dabbling with electronics. And to review teams.

Arie

No, that would remove the substantial amusement available from many scientific papers.

I saw that one with the caption, "Extra heavy-duty Russian LED".

Cheers

Phil Hobbs