What limits the theoretical efficiency of solar panels?

It doesn’t seem like it would be the second law of thermodynamics. Doesn’t that only apply to heat engines?


The Second Law applies to absolutely everything, with no exception. Yes, including that. And that, too.

That said, the practical limits on the efficiency of solar panels are far worse than any fundamental limit, at least at present. As the technology advances, there’s no reason we couldn’t make panels that were 90+% efficient.

When they talk about efficiency, are they talking about the entire solar spectrum. The work function would seem to put a major limit on things.

The Shockley-Queisser limit, stemming from the broad spectrum of solar radiation combined with the band-gap nature of semi-conductors, says that a single silicon (or similar) layer can’t beat 30%. The theoretical limit using multiple layers is still not great, at 68%. (It’s also far from realized. Multi-layer panels have only reached about 40% in the lab, and only then when using optical concentrators.)

The easiest way to take a step up in efficiency is to use solar energy as a heat source rather than an electricity source (e.g., for heating water), which can approach 70% efficiency with current technology. Trouble is, electrical energy is just so much more handy than heat energy.

Would it be possible to do better than that with some other, non-silicon material with a different band gap, though? Or if the problem is the broad spectrum of sunlight, to break it up with a prism or similar, first, and then use a collection of cells fine-tuned to the various wavelengths?

Yes, sort of. Indeed, the numbers I gave were for silicon semiconductors, and other materials have different band gaps. But, you lose in both directions. A higher band gap means the lower end of the frequency spectrum is useless. A lower band gap means the higher end of the spectrum is not being used efficiently, as any energy gain beyond [symbol]D[/symbol]E is lost to the lattice as heat (although folks are trying clever ways to recover this).

The multi-layer approach is sort of like your prism idea. The first layer catches the highest energy photons and is (roughly) transparent to lower energy ones. The next layer is tuned to have a smaller band gap, and the next smaller still. A prism itself could work, but it would only gain you efficiency if you can package the whole prism+reflectors+receivers unit under the same surface area as an equivalent “simple” panel, as the relevant efficiency is taken per unit area of panel.

I wonder if the van hoff singularities of carbon nanotubes could allow both low and high energy phonons to be formed in the same material.

Quantum dot solar cell tech have a theoretical limit of around 65% which is more than more than double the Shockley–Queisser limit. Don’t know how close they are to marketing QD solar cells, but there’s an awful lot of research being done on them right now.

I think that this technology is straddling the border between academia and industry – still lots of research groups doing research, but also a lot of start-ups and industry research trying to make something commercially viable. I know a guy who works for a company that makes various products with quantum dots, and they’ve got a few small pilot projects.

These guys who are working on oil-producing algae-do they have a real chance with this process?
I would expect that the efficiency of such processes is pretty low…

0.5%-2% are the sorts of numbers I’ve seen for algae-produced biofuels.

Photosynthetic efficiency (how well sunlight is captured in useful chemical bonds) is pretty low, just one or two percent even for most crop plants. That doesn’t necessarily mean that algae based biofuels can’t be economical. Technologies aren’t competing on how much energy they can capture per unit area, but how much energy they provide for a given cost. So, you might need (say) twenty times as much area for algae pools as you would for an equivalent solar panel installation. But solar panels need expensive crystalline silicon and non-trivial manufacturing, so they’re pretty expensive. Algae growth isn’t perfected yet, but conceivably all you need are some pools or bags of water and some trace nutrients, so it *could *well end up as a more economical choice.