Low-droop GaN LEDs in the labs...

[Benson]

A new development in GaN LEDs
A research group at UCSB led by Shuji Nakamura (the inventor of the GaN blue LED) has used a new substrate to grow GaN with a different orientation; a blue LED using this technique achieved a peak efficiency of 52% at 20mA (not sure the die size, so I can't say the current density), but the main thing is that it only drooped 14% at 10x the current. Typical white power LEDs (based on a GaN blue die with a yellow phosphor) achieve peak efficiency around 35 mA/mm^2, and show about 25-30% droop at 350 mA/mm^2. (Figures estimated from jtr1962's LED testing graphs; while the phosphor efficiency also declines at high current density, I believe this effect to be small at these values.)

If droop continues to scale as ~half the droop of existing LEDs out to >1A/mm^2, it could yield as much as 50% higher surface brightness, which as we all know means 50% more throw , plus a significant decrease in waste heat.

Of course, it'll be a while until this tech hits production, but I can wait... if only because I kinda have to...

[saabluster]

Somebody got you beat on this one already. By about three months. Where ya Ben Son?

[Benson]

Not saying you're wrong, but... You sure it's the same thing? They're both using GaN subtrates, but the older story is about a different group, and says nothing about the semipolar lattice alignment. The one I linked is being presented tomorrow @ CLEO...

I was unaware of the other story, or I'd have given more detail to clarify the difference (or what I think's a difference, anyway); as it was, I didn't mention the semipolar alignment specifically, because I'm wary of paraphrasing the bits I don't fully understand.

[saabluster]

Not saying you're wrong, but... You sure it's the same thing? They're both using GaN subtrates, but the older story is about a different group, and says nothing about the semipolar lattice alignment. The one I linked is being presented tomorrow @ CLEO...

I was unaware of the other story, or I'd have given more detail to clarify the difference (or what I think's a difference, anyway); as it was, I didn't mention the semipolar alignment specifically, because I'm wary of paraphrasing the bits I don't fully understand.

Hmm. Maybe that is different. Don't know. I just saw Shuji Nakamura, reduced droop, and GaN substrates which are of course everything they talk about with Shuji's company Soraa. Buuut...tilted crystals? Don't remember reading anything about that one so maybe that is indeed a new development or it could just be further explanation for what they are doing with Soraa or even similar tech but applied in the laser field. That picture they show on the site looks more like a laser diode than an LED. We shall see I guess. Make sure to post with more info once that becomes available.

[bshanahan14rulz]

I read about this several months ago as well.

The problem is that it is very hard to grow crystals large enough in the semipolar arrangement, with few dopant errors, and have a large enough plane to cut chips out of.

Just like how it was a problem to find a suitable material with the correct bandgap to make our shorter wavelength blue LEDs years ago.