An LED, or a light emitting diode, is based on the emission of light by a semiconductor. Of course, there are many semiconducting materials out there and what determines the color of the light they emit is something called their bandgap. The bandgap is the energy difference between the highest occupied set of electronic levels (the valence band) and the next set of levels up (called the conduction band). The way an LED works is that you use electricity to kick up an electron from the valence band to the conduction band. As the electron then relaxes back down, it will release (some) energy as light. The wavelength of the light will then depend on the bandgap, . The key rule is that larger bandgaps equal bluer light and smaller bandgaps give you redder light.
Now the way temperature comes into play is that it affects the spacing between the atoms, and by extension the bandgap. A simple explanation is that higher temperatures cause atoms to jiggle more, which causes them to be further apart from each other. It is for this reason that crystals generally expand when it's hot and contract when it's cold. Now in most cases when you bring atoms closer together the bandgap also increases. This effect has to do with the stronger interaction between the electrons in this tighter mesh of atoms. As a result, the color of the light emitted shifts towards the blue end of the spectrum. You can nicely see that effect in this GIF. As the LED cools and the bandgap increases, the light goes from orange-red to yellow, and finally to green.
Source for the GIF: this video
Sure! I am a bit surprised that you saw no effect whatsoever, but maybe you were just a bit unlucky in your selection of LEDs. It is certainly true that different materials will have a different temperature dependence. For many semiconductors you can approximate the temperature dependence of the bandgap by:
E(T) = E(0) - aT2/(T+b),
where a and b are empirical constants. The equation above is called the Varshni equation and is described in more detail in this paper. In the same paper you will find values for a and b fore some common materials used in LEDs. As you can see there is quite a bit variation in these values, meaning that the change in the bandgap (and color) will be more dramatic for some materials than for others. In fact, in rare cases you will see the bandgap decrease as the temperature is decreased over a certain range. Off the top of my head, I know this is somewhat common in some lead salts like PbS. Finally, the Varshni equation only works well for typical inorganic semiconductors. If you use LEDs with an organic active material (OLEDs) or other classes of materials, the effect could be different.
Super simple, just need some liquid nitrogen.
Works for people, too.
Thanks for the explanation! I had to test some LEDs in liquid nitrogen recently as part of a project at University. These colour changing effects didnt occur at all with any of the LEDs I used. Why would this be? Different semiconductor materials that aren't as responsive to temperature changes?
Damn it! At first I wrote by "simply changing the temperature." Then I wanted to make it more specific by mentioning the liquid nitrogen, and ended up with this mess. I shouldn't have posted before drinking my second morning coffee...
The only thing I can remember changing was intensity and some of them (of course) failed if dipped into the liquid nitrogen bucket too quickly. Thanks for your reply and link to the paper, I'll check it out! Might need it for my Masters anyway!
Yeah let me grab some from my garage real quick brb