On the subway home tonight, I was pondering…
It’s not too difficult (though not easy) to make an incandescent light bulb. You blow the bulb, fit it to the base and wires, pump out the air, and seal it off. I can imagine a determined amateur glassblower with ample reources gaining the additional skills to make the base, bond glass to metal to seal everything off, then pump out the air and seal that opening… especially if he or she already has descriptions of how it is done.
Given those skills, I can further imagine the hobbyist making fluorescent lamps: they involve the same glass tubing, endcaps, and wires; the additions are the fluorescent coating inside the glass, and the mercury or whatever it is that goes inside the tube.
Light-emitting diodes (LEDs), though, would seem to be a thing of an entirely-different order of difficulty. They involve ultra-pure semiconductors and fine-scale construction techniques.
Have any hobbyists ever made their own LEDs?
I’m not going to say it is impossible for a hobbyist to do, but it sure ain’t gonna be easy.
You’re not looking to make computer chips, so a piece of dust here and there probably isn’t going to be catastrophic. After all, you just need one LED. It doesn’t matter if you make a hundred DEDs (dark emitting diodes, a somewhat humerous way to refer to a dead LED) along with it. By comparison, if you are making a computer chip, one little piece of dust just killed a few megs of ram. Dust is a lot more important when you really go miniature.
The trick here is that semiconductors aren’t made, they are grown. When I was little, I made candy by mixing sugar in water, then placing a string (tied to a pencil) in the water. Every day, lift up a bit on the string (by turning the pencil) and soon you have a nice glob of crystalized sugar on the end of your string. Growing semiconductors is the same basic idea, only instead of sugar water you have some molten metal like silicon. It is typically grown into long crystal tubes, then sliced up into wafers. The trick to making a semiconductor is that you accidentally muck up your crystal while you are growing it, but you muck it up in a very specific way. You need to get a fairly even layer of very specific impurities in there, like aluminum, gallium, arsenic, or phosphorus (or a few others). Exactly what impurity you use gives the growing crystal its semiconductor properties. The exact combination of metals and impurities that you choose will determine what type of semiconductor you end up with, and in the case of an LED, what color you end up with as well. Gallium, phosphorus, aluminum, and arsenic tend to be common in LEDs.
Construction techniques vary, but probably the easiest way for a hobbyist to do this is to grow his/her semiconductor crystal in a vacuum chamber and injecting vaporized impurities into the chamber while the crystal is growing. It’s probably not something you could do in a couple of hours over your back yard grill while sipping on a Bud Light, but it’s probably not completely above the abilities of a really motivated hobbyist either.
A hobbyist no. The equipment would be too expensive for a hobbyist, and it’s no longer a hobby it’s a manufacturing plant when you buy all the equipment.
Seems to me that it would probably be possible for a hobbyist to make some kind of LED, provided that it was acceptable for it to be horribly inefficient and very low output.
In fact, it could be quite simple; all semiconductor diodes emit light, and you can make a semiconductor diode out of naturally-occuring minerals - this is how a crystal radio works. I expect the kind of cats-whisker diode in a crystal set would only produce a very low output, probably in the infrared, but it ought to be possible to make a crude visible (but dim) LED by similar methods.
it is probably feasible for a hobbiest to make an organic LED - the chemicals needed would be available through a chemical supplier. You would then need ITO covered glass which is also probably available. A metal evaporator is perhaps the most tricky thing, but they are available second hand for a few thousand.
However, the devices tend to only last a minute of two, but are neat. Used to make them as a post-doc.
e.g. see people.ccmr.cornell.edu/~cober/ mse542/page2/files/Kochanek%20OLEDs.pdf for an advanced primer
Checkout the bottom half of this page.