I don't see a "future technologies" forum, so I guess this is probably the next best place for this.

Noticed an interesting article in IEEE Spectrum about using radioactive radiation instead of solar radiation for power generation in "solar" cells. Interesting article... Had me curious about how much power said cells could generate in space, perhaps near the Sun or Jupiter.

Here is the Spectrum article

http://spectrum.ieee.org/feb07/4887

After much searching, here is the USPTO patent application

http://appft1.uspto.gov/netacgi/nph-Parser...ancis+AND+Tsang

If it works it could turn burned-out reactor fuel elements into an asset rather than a liability

These seem like a good idea but to date the trade off's are still better going the RTG route (where you capture the radioactive energy in some absorber thereby providing a heat source and then use a Peltier type junction to generate electrical power from a temperature differential) rather than a direct RF energy conversion via semiconductor (as Photelectric cells do for visible RF radiation). These trade offs are not just direct conversion efficiency but things like the practicality of shielding (Pu-238 is fairly efficient for shielding effectively for example since most of the radiation is in the form of alpha particles) and material degradation have to be factored in (high energy RF radiation has a tendency to transform the delicate semi conductor junction structures the conversion depends on).

Heres a completely un validated source that has some intetresting information. The fact that we have yet to see a viable non thermal radioactive power source is quite telling IMO.

Helvick,
Oh, I agree, but it does offer some intriguing possibilities. The efficiencies aren't there yet, 1% is what they quote, but solar cells started really inefficient as well, and are now in the ~30-40% range. RTG efficiency is only in the 3-7% range, so there isn't an insurmountable hurdle to reach. Plus, there's no reason not to piggyback the two technologies, the cells get first bite, surrounded by the RTG thermocouples.

Also, an RTG can only get so small, but these cells could be made arbitrarily small, or large. Might work well for a penetrator that needs a small long term power source.

As for shielding, I assume the cells themselves would act as shields around the power source. You'd need extra protection to prevent a release during a launch failure.

One nice aspect of the design discussed is that it uses liquid semiconductors, so there is no degradation of the junctions; they are self-healing.

Like you said, it's probably a technology still years away from usefulness, but hey, we can say the same thing about fusion...