3 ms·
Disappointingly this article doesn't provide a link to the research. I think this paper might be it[0], but I am not sure. So in the paper they develop a materi
by gene-h 7y ago
Disappointingly this article doesn't provide a link to the research. I think this paper might be it[0], but I am not sure. So in the paper they develop a material that has an estimated storage efficiency of 3.8%(this might be solar conversion efficiency) and energy density of 0.48 MJ/kg. This material is what's called a photochromic meaning it changes color when exposed to sunlight. This is just like those glasses which darken when you go outside. Well it turns out that many photochromic molecules accomplish this by changing or reconfiguring their molecular structure, with azobenzene being the classic example of this[1].
Now as you may know color changing glasses eventually change back to their original color, this is because the color changed state is at a little bit higher energy than the clear state. There might be a bit of a hill it has to overcome before it can go all the way downhill to the lower energy state, but eventually random thermal motion will push it over. So because the color changed state is at higher energy it can release energy as it falls down.
So the interesting thing they show here is that their molecule has a half life or rate at which it converts back to its original form of 10 months, which is a pretty long time. The energy density is also pretty good too, 0.48 MJ/kg is about the energy density of alkaline batteries. And while 3.8% energy conversion efficiency seems pretty bad it's pretty good for photochromic materials of which efficiencies less than 1% are typical. Of course they want to put this in windows and since you can't make the whole window out of this material the energy density of the material+window plastic ends up being ~2.7-3.8 kJ/kg. Of course there's another problem here too. The material they develop is yellow in its uncharged state only converting to clear when exposed to sufficient sunlight, which may not be very attractive to consumers.
[0]https://research.chalmers.se/publication/510123/file/510123_Fulltext.pdf https://research.chalmers.se/publication/510123/file/510123_...
[1]https://en.wikipedia.org/wiki/Azobenzene https://en.wikipedia.org/wiki/Azobenzene
[2]https://en.wikipedia.org/wiki/Energy_density https://en.wikipedia.org/wiki/Energy_density
- DoctorOetker 7y agotable 1 of the paper you reference [0], also lists "N5d" with a half-life of 18 years I am confused with the different efficiency metrics, one kind of metric hovers around the single digit percentages, but Table 1 lists Φ/photo conversion [%] between 46% and 77% ... This is substance 2l in the referenced paper at https://doi.org/10.1002/chem.201802932 https://doi.org/10.1002/chem.201802932
- gene-h 7y agoThe single digit efficiency metric is probably overall conversion efficiency of sunlight. The two efficiencies are probably quantum yield, or the efficiency at converting one wavelength of light. We have two because the molecule's tristable. Converting from state 0 to 1 has one efficiency while converting from state 1 to 2 has another. State 0 and 1 also absorb different wavelengths of light. Because sunlight's a mix of different wavelengths not just the two our molecule's active at, the efficiency at sunlight conversion will be different than the quantum efficiency.