CREE
First I would encourage you to go to this message and try your best to follow the first link:
Message 12785409
It is technical but many slide state a conclusion or have good enough graphics that may help. Don't try to understand the physics behind thing like figure of merit but rather that materials like GaN and SiC will be greatly superior to current device materials (GaAs, Si). That link won't answer your question in full necessarily but may help you to better understand the following.
1. First, I understand they have a proprietary technology to grow SiC. I don't understand to what extent, as you and Mike said, that this proprietary technology "allows them to use the raw SiC materials to make specific devices."
Not only do they have proprietary technology to grow the fundamental building block (ie growing SiC wafers) they possess proprietary technology for taking that raw material and creating devices out of them. The fore mentioned link has cross sectional drawings of how some of the devices are made. To make such devices requires doing certain processes like: doping (changing the electrical nature by adding impurities which give a region a net positive or negative charge), applying a layer (this can be an insulative layer to inhibit electron flow, a barrier layer to prevent intermixing of materials or chemical reactions between them, a buffer layer to improve the fit of different crystals to one another, or finally an active layer; this is sometimes done in place of trying to dope a region), and finally oxidation which allows one to grow (as opposed to applying a layer)an insulating region. CREE has patents and certainly trade secrets that relate to all of the above processes with the possible exception of doping.
Their efforts have allowed them to first find and apply the methods that are most suitable for practical manufacturing of devices built upon or into the SiC substrates. I infer this since nobody else seems to be commercially active in the SiC game other than possibly Westinghouse and Nitres. Furthermore they have patented processes which allow reworking defective materials. This is a big plus when trying to control costs.
Now add on top of the processing techniques that they have patented the structure of some of the devices themselves. That is sort of like owning the source code or blue print. BTW it is also a lot easier to see if someone has violated these types of patents than the process patents.
I hope this makes it clearer that CREE has IP in three different areas regarding SiC: growing the stuff; processing the already grown stuff; final devices grown out of the stuff.
2. I understand they, along with others, make blue LEDs. I don't understand the proprietary nature of this market or how, as you said, Cree's IPR in this area forces other manufacturers' blue LEDs to be "mostly made in cost disadvantaged material/process."
Part of this is answered by the above, but the other thing to realize is that there are two flavors of blue LEDs: those grown on a sapphire substrate and those grown on a SiC substrate. The advantages of SiC have been explained numerous times by unclewest and jacobs. I try to state it slightly differently. Assume the devices have all the same processes at the same cost and the substrates cost exactly the same. The SiC being conductive allow itself to an electrical contact. Sapphire does not; therefore a relatively large area on the device top surface must be allocated for the second contact point. Ergo, the sapphire devices are much larger. That means fewer parts per wafer at the same processing costs. The sapphire based parts will be price disadvantaged.
Finally Frank, sorry to post this esoteric stuff here, but the question was asked on this thread so I thought it best to answer here.
FATBOY |
