Germ Anium
Intriguing, yet frustrating in its lack of content, this press release disguised as an article in Red Herring raises more questions than it answers.
While the high-efficiency material is expensive, SolFocus' system uses so little of the germanium — 1 square centimeter per panel — that the company is able to cut the installed price of solar in half, compared to the average price today, and still make a healthy margin.Gives us the size of the panel.
Describe the geometry of electrodes that the wafer uses.
You need a lot of current to overcome the low band-gap of germanium.
Update: For more info look at the comments below; PV-expert SW provides the needed clarification regarding the technology. As more info becomes available, I will try to summarize in full. In short, I may have to change this post's title from Germ Anium to Gallium Arsenide. GaAs has double the bandgap of Ge, so at least that question gets answered.
posted by @whut at March 24, 2006
8 Comments:
All these guys make are the optics. The cells in question are either GaAs or Multi-junction GaInP/GaAs/Ge. Most III-V solar cells grown via MOCVD sold commercially for the space power industry are grown on Ge substrates because of the close match between the lattice constant of Ge and GaAs. SpectroLab and Emcore are the primary sources of the cells currently. These are the cells that (together with the dust devils) are responsible for the incredible lifespan of the Mars rovers. The idea of using concentrating optics is an effort to find a way to make these devices economically viable for terrestrial power generation because of their high cost per unit area.
To follow up on that, this looks like a piece of decent technology and poor journalism.
Notice these guys are realistic. They aren't claiming that they are "revolutionizing PV" making it dirt cheap or anything. They are simply trying to find a way to use relatively cheap optics and these very expensive but highly efficient cells grown on Ge substrates to lower the cost of PV generated electricity. They figure that in areas of high insolation (where most of the spectrum is direct most of the time) they can cut this cost in half.
I think they are probably correct.
SW, Do you really think they can go through the bulk manufacturing process on Ge? I see huge advantages with space applications where you have one-off circuitry and can afford just about any robust solar-powered module.
Yes, I realize the close lattice match between GaAs and Ge, but they also have the problems with anti-phase domains because of the polarized GaAs compound. These of course lead to defects & traps and reduced efficiency. And the lattice match is not perfect so with increasing thickness of the epitaxial layer, the strain will monotonically increase, leading to potential for more defects, and possible breakdown.
Of course, the potential is there to make solar cells with lots of defects, think amorphous silicon, but I find it interesting that they are going this route.
Thanks for the tip, I did not realize that the GaAs was the key point in the technology.
And I agree totally that the journalism was completely lacking. That's why I asked the questions, I tried but couldn't find any other references online.
GaAs on Ge is notable for its singular success, to my knowledge in the hetero-epitaxial business. High efficiency GaAs cells grown on Ge substrates approach the performance of those grown on GaAs substrates. One of the truly wonderful things about operating these diodes under concentrated light is that what few defects and recombination centers remain in these crystals have a way of saturating out, improving their performance as the intensity increases. Of course you have to deal with resistive losses and thermal issues but the cells themselves, the current voltage characteristics actually improve often dramatically.
Here is a page at Spectrolab that has some cartoons of possible cell structures. Spectrolab
Except if they get used as laser diodes or (less extreme) LED's in which case the defects can cause them to essentially self-destruct. Interesting to know that the reverse process, i.e. photons to electrons, saturates before hitting a thermal runaway point.
Moreover, it's crackers to slip a rozer the dropsy in snide.
Sure, J.D., sure.
Post a Comment
<< Home