Why are diode "projections" rectangular?

[Traveller]

I'm sorry if this is not exactly a question on optics, but why do diodes project rectangular "background" light, as well as vertical lines, etc.?

Using a highly efficient lens from JayRob, (as well as Merediths, or so I've heard), these native traits get amplified. Below, a perfect example of a 405nm 6X diode projected onto matte-black paper as well as white grid paper. The diode used is also shown.

The first image is with a typical AixiZ lens as thus doesn't show off the byproducts but said byproducts are visible with 400-700nm safety goggles (maybe associated with UV light, dunno).

Thanks to the crew for any info (and sorry if this has already been covered elsewhere, but I didn't find anything).

 

[daguin]

The surfaces of the die (crystal) are not "round." This leads to light "leaking" around the entire front surface and small amount leaving the crystal "less than coherent." Also, there are other "components" in the crystal that "can" lase at a different wavelengths included on the structure of the crystal. Different wavelengths colimate at different focus lengths, so that light would be unfocused at the distance that you have the primary wavelength focused.

Your last photo in the montage shows these phenomenon quite nicely. You are "seeing" the actual shape of the front of the die and the pattern of non-coherent light"fanning out" from the primary lase point.

Peace,
dave

 

[Traveller]

... You are "seeing" the actual shape of the front of the die and the pattern of non-coherent light"fanning out" from the primary lase point.

Hey Professor,

that really helps to build a better picture   I read up on DPSS constructions due to the IR leakage but I didn't realize that with one-crystal solutions such as our red & BR diodes, we could experience something "similar" in nature.

Thx for "stopping by"  

 

[GooeyGus]

or you can imagine how a HeNe laser glows. It still gives off light that is not coherent. Same thing happens in the tiny die of a laser diode, you get light from the die being energized that isn't necessarily coherent, just like a HeNe throws off light that isn't coherent from the gas being energized

 

[pullbangdead]

Yep, you have some pretty good images there of exactly what the die looks like. As the diode is made, you have a relatively thick substrate, and then the layers are grown on top of that substrate. All the layers pu together still add up to only a microscopic thickness on top of the substrate, so you can see it looks like all the light comes from the very top surface of the die there (eg, your last picture is a good image of the front facet of the die, but it's actually "upside down" from how the thing is built). The layers doing most (all in an ideal case) of the light emitting, called active layers (real creative, eh?), are actually quantum wells: only single-digit nanometers thick each, but there may be from 1 to 5 or so of them, each separated by another layer that is of equal thickness.

A typical diode may have something like 3 or 5 wells, each 5nm or so thick, each separated by a 5nm thick or so layer. Ideally, electrons are flowing up through the diode (from just above the substrate, but not through it, sapphire isn't a conductor), and holes are flowing downward from the top of the diode. There generally are some layers above and below the active layers that trap the electrons and the holes in the active region. The electrons and holes then recombine in the active layers, and emit light when they do so. In an LED, it's incoherent light. But in a laser diode, you get stimulated emission instead of spontaneous, since you have feedback from the facets, and you therefore get coherent light. But there is still spontaneous emission occurring as well, it's just tiny compared to the output of the stimulated emission. You generally can assume the LED incoherent light from the diode to be zero once lasing takes over, but there's still *some* there, so you get some LED-style light from the die.

Or that's the ideal case. But you end up with carrier leakage, holes and electrons that go places they shouldn't, and recombine in places other than in the active region, so you get a little LED action in lots of places, and lots of heat (these recombinations outside the active region can generate heat instead of light). And LED-style incoherent light is emitted in all directions, randomly, so it can come out of the die anywhere, in any direction, even if it is emitted from the active region. And that's where some of your inefficiencies come from, in addition to simple resistive heating from the current flowing through the die.

 

[Traveller]

Wow... thank you, that's really quite interesting how the diodes are "constructed"! It certainly explains what we're seeing, that's for sure  

Thanks for taking the time to explain it in such detail - much appreciated  

 

[maxh]

On a similar topic, I've noticed that some LDs have oval shaped outputs, such at a phr diode. I've been assuming that's due to the type of construction as illustrated in the following picture (from this website: http://dictionary.zdnet.com/definition/laser+diode.html)

So are the typical blu-rays the gain-guided stripe kind?

 

[dr-ebert]

The picture is from 2000; there's a good chance that existing diodes will deviate significantly from these types...

 

[pullbangdead]

It is due to the type of construction, and those shown are popular types of diodes, that are certainly still used in research and possibly in production.

As to what the structure of any given diode is, it's impossible to say without knowing how it was grown and processed, or without reverse engineering it (which is NOT trivial). It could be gain guided, it could be what is commonly referred to as a ridge type laser, it could be any number of more complicated designs. It's not a VCSEL, all the laser diodes we have from optical storage drives are edge emitters, but beyond that it's very difficult to ever tell exactly what kind of structure they used when processing the laser.

These diagrams are also not necessarily completely correct, they're just schematic, so take that for what it's worth.