Lower Priced Silicon Possible with SRI Process

Red Herring reports that SRI International has licensed technology to produce lower cost solar-grade silicon to three Asian companies and that pilot plants could be up and running in 18 months.

SRI’s technology promises to make solar-grade silicon for $14 per kilogram, less than half the price of competing technologies, Mr. Dubois says. SRI claims its process produces silicon with almost no impurities (0.02 parts per million) at one-tenth the energy and one-fifth the capital costs of the conventional Siemens process used by most of the big polysilicon manufacturers. So far, SRI’s technology has only been used in the lab. But analysts say that, with contract prices rising up to $85 per kilogram and non-contract prices skyrocketing to $200 per kilogram, companies are willing to try new technologies.

A 1986 patent by A. Sanjurjo, assigned to SRI, is described as follows:

Apparatus is described for producing low cost, high purity solar grade silicon ingots in single crystal or quasi single crystal ingot form in a substantially continuous operation in a two stage reactor starting with sodium fluosilicate and a metal more electropositive than silicon (preferably sodium) in separate compartments having easy vapor transport there between and thermally decomposing the sodium fluosilicate to cause formation of substantially pure silicon and a metal fluoride which may be continuously separated in the melt and silicon may be directly and continuously cast from the melt.

Comments

[Chad]

If we can just find out what solar companies they're selling this to... of the 3 that own their own polySI production... I'd know who's stock to buy [more of] :)

[Greg woulf]

If the patent is legit, and there aren't other patents blocking something vital then anyone can do it, I think.

I'm not a patent attorney, but 20 years is the life of a patent.

In any case, this is fantastic news. That should translate into half price panels or so sooner or later.

[John F.]

Here is the patent, in case you want to read it yourself.

I'm not a patent attorney, but 20 years is the life of a patent.

I thought it was 17 years.

I can think of two possible explanations, either or both (or neither) of which may be true:

1. The claims of superior economics are exaggerated.

2. They've made recent key improvements to the technology that both extend the IP and make the process workable, whereas it wasn't before.

The latter explanation has some support in the Red Herring article, where it mentions that the SRI scientists have "updated" the technology.

[David Sarokin]

According to some newly released research, PV solar panel use in homes is due to soar in the next five years:

The Growth of Photovoltaic Solar Energy for Home Use

But you all knew that, didn't you!

David

[John F.]

Actually, patents used to be in effect for 17 years from the date of issue. Starting 1995, it changeed to 20 years from date of filing. But that works out to be about the same thing, since there's usually a 2 or 3 year interval between filing and issue.

[buy viagra]

Hi, nice post. I have been wondering about this topic,so thanks for sharing. I will certainly be subscribing to your blog

[Term Papers]

But analysts say that, with contract prices rising up to $85 per kilogram and non-contract prices skyrocketing to $200 per kilogram, companies are willing to try new technologies.

[Furniture Stores in Los Angeles]

Feels like more and more things are using silicon. Lower prices is a great thing.