Physics Stack Exchange is a question and answer site for active researchers, academics and students of physics. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I don't know much about light-emitting diodes, but I imaging if you had a panel of RGB diodes you could produce any wavelength of color within the visible light spectrum. However, if I also wanted to generate specific wavelengths of UVA or UVB (anywhere from 290 to 400nm), could I also accomplish this using diodes? Essentially, I am interested in making a small panel of diodes in which I could produce any specific wavelength of visible light, UVA, or UVB.

Thanks in advance.

share|cite|improve this question
Not in the least germane, but...a cheap UV LED and a mason jar full of liquid stain remover (some brands fluoresce) were all it took to make my Jack-o-lantern glow evil blue this year. For next year, several LEDs on a board for more even coverage, and a pulsing supply... – dmckee Jan 6 '11 at 4:43
@dmckee: I'll have to try that! – Mike Dunlavey Oct 19 '11 at 11:27
@dmckee Which brand works best? – Alex1167623 Apr 7 '12 at 23:52
@Alex1167623 The only one I've tried is OxyClean. use the "Free" stuff, no dye, no scent. Since then I've also learned that the quinine in tonic water will fluoresce nicely as well, though it generates a whiter light. Try them both if you can. – dmckee Apr 8 '12 at 0:13
@dmckee THX. Oxyclean is not sold in my country but I will try Tonic water. I will also solvate the refill package ink of a text highlighter in water and give it a try. – Alex1167623 Apr 8 '12 at 8:56

RGB diodes, the kind you find in an LED computer monitor or television, don't actually produce any wavelength of light you want. As you probably know, the diodes come in triplets, one each of red, green, and blue. Each of those individual diodes only emits one specific wavelength of light.

However, by varying the relative intensities of the three colors, it's possible to simulate the effect of any other wavelength of light on the human eye. Since there are only three types of cones on the retina, for any given wavelength which the eye detects, you can induce the same chemical reaction in the cones as that wavelength naturally would, using only three colors.

The point is, in RGB LEDs there are only three wavelengths ever being produced, and it's only by "fooling" our eye that they create the illusion of the full spectrum. Since human eyes aren't sensitive to UV radiation, the trick wouldn't work for ultraviolet wavelengths. You can't actually create a diode that will produce any wavelength you want dynamically.

Of course, there are a wide range of materials that you can put in a diode in order to affect the wavelength of radiation it does produce. From a physics perspective, UV radiation is no different from light except that it has a higher frequency (shorter wavelength), so by choosing the right material, you can in principle create a diode that will get you any pre-chosen UV wavelength you want. The Wikipedia article on LEDs gives a table of various materials grouped by the wavelengths they can produce, and there are several options for UV radiation. (There may be a lower limit on the wavelength obtainable from an LED due to energy considerations, but I'm not familiar with the details of how LEDs work so I couldn't tell you whether that's the case or what wavelength it might be.)

share|cite|improve this answer
""The point is, in RGB LEDs there are only three wavelengths ever being produced,"" Have a look in Wikipedia LEDs! Each LED produces a rather broad band, far from "a wavelength". – Georg Apr 25 '11 at 18:28
@David: Something I've been wondering. If you mix red and blue light you get a color called "magenta" - kind of a light reddish-violet color. Am I right that this color does not exist in the rainbow (because red and blue are at opposite ends), and it's just the label our brain gives to that combination of stimuli? – Mike Dunlavey Oct 19 '11 at 11:21
@MikeDunlavey Exactly right. – Kevin Reid Dec 5 '11 at 15:31

Light-emitting diodes (LEDs) can be manufactured to emit light in the ultraviolet range, although practical LED arrays are very limited below 365 nm. LED efficiency at 365 nm is approx 5-8%, whereas efficiency at 395 nm is closer to 20%, and power outputs at these longer UV wavelengths are also better.

There are not much options in production currently. Mainly ~395nm is available and that's it. Spectrum is quite wide. You can change wavelength a little by altering current/temperature, but that's tricky.

You might want to look at non-linear crystals too (KTP for example).

share|cite|improve this answer
Now 365nm LEDs are quite common, and even shorter-wavelength ones are available though very expensive. – BarsMonster Feb 15 '15 at 15:57

Leonidas is completely wrong! The color of light produced from a light emitting diode is manly dependent of the band gap of the semiconductor material that is used to produce the LED.

For example: • Infrared: GaAs or AlGaAs • Red: GaInAlP or GaAlP • Yellow/Orange: GaInAlP or GaAlP. • Green: GaAlP or InGaN • Blue: GaN or InGaN. • Ultraviolet: GaN.

You may note that some of the materials are the same for different colors of light. This is due to the change in band gap that can be produced by changing the doping concentration and doping materials.

Some LEDs have a layer of fluorescent material though. To shift the color slightly in an easy way.

And to answer the original question: Yes, as BarsMonster pointed out, there are Ultraviolet LEDs, but the selection is not that big.

share|cite|improve this answer
That used to be a method for producing white LEDs - a blue LED and a fluorescent dye. It's not used anymore because of true white LEDs, and being able to package multiple wavelength LEDs on the same die. – Martin Beckett Apr 25 '11 at 16:25
@Einar For me being completely wrong you seem to admit me being right a lot in your fourth paragraph. OP asked about LED-array with shiftable wavelength in UV. – Leonidas Aug 7 '11 at 22:55

RGB diodes are actually LED with fluorescent dye which is stimulated and thus mixed to different colours. LED-light itself is nearly monochrome.

So unless you find LEDs with dye that fluoresces in UV (I doubt that there are some for the normal market) you'll have to buy some UV-LED for these wavelengths.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.