 |
| Energy diagram of a dye-sensitized photovoltaic cell, from this paper. |
This is a compelling idea, and a main selling point is the hope that devices based on this could be cheap and much less energy-intensive in their manufacturing, since they could be made with materials that don't require high temperature synthesis or high purity, like Si solar cells. After many years of effort, end-to-end power conversion efficiencies are up around 13% for outdoor solar illumination-type conditions. Here (link to NIH free version) is a good review from 2021 that is very complete in summarizing progress.
So, 13% is nice, but it's hard to see that being competitive with Si for bulk photovoltaics, and perovskites, also solution-processable, are up over 25%, similar to Si. Still, outdoor solar is not the only application! This paper from 2017 showed that it is possible for dye-sensitized cells to get power conversion efficiencies up to around 30% for indoor lighting conditions (much lower intensity, different spectra than solar illumination). That seems to be the basis for this story in today's Wall Street Journal, pointing out that there are actual consumer products coming to market that have dye-sensitized cells for indoor operations. Very cool to see this product really start to make it out of the lab! If one of my readers has a good, clear explanation of why the power conversion efficiency is so much better for indoor lighting conditions, please leave it in the comments.
Indoor light from daylight coming in or indoor light from illumination by electric lighting...?
ReplyDeleteCould the different indoor spectrum (different (both) from the (atmosphere filtered) solar spectrum) be the cause? I.e. a better matching to the absorption of the cells?
Indoor lighting from electric illumination (fluorescents or LEDs, or perhaps incandescents, though that seems unlikely these days).
ReplyDeleteOkay, thanks.
ReplyDeleteThat does make me wonder what the purpose is. Transition from electrical energy to light and back to electrical energy (the latter step with an efficiency of 13%...) seems a waste if either one can plug in or use wireless charging (that I think has a lower loss)?
It's an interesting result, which I had missed. Thanks for pointing it out. Any chance you have a non-paywalled link to the WSJ article?
ReplyDeleteOrdinarily, efficiencies of solar cells go down as the illumination intensity decreases (for thermodynamic reasons, related to entropy, at root), so it's funny to have such a high power conversion efficiency (PCE) for low-intensity indoor light.
But the fluorescent tubes they are using do not produce thermal radiation, and that's a big advantage. With a black-body spectrum, when you choose the absorber material in your PV cell, you have to compromise between lower bandgap (letting you absorb more of the incident spectrum, giving high current J) and higher bandgap (giving higher extracted energy per electron, higher V). Since power is J*V, every solar cell is a compromise. But the fluorescent bulbs have about 6 sizeable emission peaks from 350 to about 700 nm (Figure S5). This non-thermal source is easier to convert to electricity, as you can choose a material with a bandgap near 650 or 700 nm, absorb all the light, and still get a decently high voltage, if the device works well. (As an aside, in the more extreme limit of laser power converters, device efficiencies over 70% have been demonstrated, and the theoretical limit is 100% at high laser power). Their device clearly works well in this range, with an absorption edge near 700 nm and very efficient conversion of wavelengths below 650 nm (Fig 3).
So, in this case, the DSSC normally has an effective band gap that is too large for the solar spectrum, but it's really nicely chosen for this indoor lighting scenario. The paper compares the DSSC to a reference GaAs device and shows that their DSSC does a better job. That's really quite impressive, as GaAs are the best single-semiconductor solar cells around, and they used best-in-class cells from Alta Devices.
Yes, that expert answer was the root of my question (first post). Better spectral matching (one aspect being less IR). Thank you!
DeleteJ, thanks - great comment. Sorry, I can't find a non-paywalled version of the WSJ article.
ReplyDeleteTuned nano dots.
ReplyDelete