The chief technical officer of Soraa has written a paper detailing the benefits that flow from building GaN LEDs on a substrate of the same material. It's a long list.
We have touched on Soraa's GaN-on-GaN process before, but the paper penned by Mike Krames goes into far more detail and outlines, by my count, 10 advantages that Soraa claims directly result from GaN-on-GaN. Let's have a look at them.
Simplified wafer processing. LEDs on sapphire, silicon carbide, or silicon require more steps in the manufacturing process -- wafer bonding and lift-off -- and more material -- the carrier -- that are unnecessary in GaN-on-GaN.
Lower defect density. Soraa claims an advantage of a factor of 1,000 here, down from a million defects per square millimeter for a typical LED on sapphire to around a thousand for GaN-on-GaN.
Higher light extraction efficiency. This results from the match of optical refractive index between LED and substrate.
More light output density from smaller chips. Krames writes that the "total light output capability for the GaN-on-GaN LED is equivalent to the best LEDs reported, but in a chip size 15-25 times smaller."
Better heat dissipation. The uniform GaN environment means that both heat and current conductance are increased. Fewer defects mean fewer hot spots, and heat is conducted away from any such spots more quickly. Soraa's latest LEDs are rated at 120°C and can be used in enclosed fixtures.
Higher current with less droop. Krames implies, but does not say directly, that LED droop is essentially not a problem with GaN-on-GaN.
Full-spectrum color. Soraa's LEDs have their primary emission in the violet part of the spectrum, instead of the blue that characterizes almost all other LEDs on the market. This fills out the emission spectrum at both ends when compared with other phosphor-converted white LEDs: on the violet end because of the primary emission, and in the red because the higher energy of the primary emission drives deep-red phosphors effectively. The Stokes gap at cyan is also reduced.
Brighter whites. Optical brightening agents have been added for decades to fabric, clothing, and paper, which then appear brighter under sunlight or incandescent light. "Standard blue-based LEDs cannot excite the OBAs," Krames says, but Soraa's violet primary emission does.
Single-beam operation. Soraa's MR16 lamps, for instance, don't incur the "ghosting" that results from multiple LED sources in some competing products. Soraa can achieve this because their LEDs are uncommonly small and uncommonly bright, enabling a 50-Watt halogen replacement bulb using a single LED.
Potential for manufacturing savings. Since the main objective of lighting makers "is not to ship wafers, but to ship lumens," Krames writes, Soraa's "higher current density means that one lot of GaN-on-GaN wafers is equivalent to about five to 10 lots of standard LED wafers."
That is a lot of claimed advantages, and I don't see logical weaknesses in any of the claims. If GaN-on-GaN does all that Soraa says it does, it may well deserve the moniker of "the second-generation of white LEDs" that Krames suggests. Then why aren't more players in the industry investigating GaN-on-GaN?
I presume that Soraa has patented key elements of their process; but they could license it to others. Are they willing to do this? If so, has anyone taken out a license? My questions to Soraa were not answered by press time. Can anyone here share any insights?
— Keith Dawson , Editor-in-Chief, All LED Lighting