Cheap material converts heat to electricity

[Corey]

Posting for the TEG crowd out there: https://www.sciencemag.org/news/2021/08/cheap-material-converts-heat-electricity

I believe the zT 'figure of merit' they quote as ~3.1 for this new polycrystaline tin selenide material means it would be roughly three times as efficient as the current standard bismuth telluride - which I typically see quoted around 1.0. So assuming they can keep other factors relatively in check, this might be an interesting material for near-future development.

Article gets a little out in the weeds by the end saying people might hook these up to water heaters to create electricity. Well, I guess if you have a water heater where you don't actually want hot water! Otherwise, you'd be much better off to insulate the water heater and put this somewhere you actually have waste heat!

 

[peakbagger]

Its an interesting line of research but I would not start saving up my nickels to buy one of the new generation.

 

[stoveliker]

The scientific literature abounds in "high-ZT" papers, but I have yet to find the materials described in them for sale in economically useful products.
However, going from single crystal to some compressed polycrystalline material (and this not being some theoretical calculation predicting a record ZT...) does render this closer to "practical reality" than most experimental cases I've seen.

I'm not sure I want a lot of selenium around in my home though, and (unless used on a small scale in electronics that already contain a zoo of not so healthy elements) for "human-scale" (as opposed to microelectronics-sized) applications one would need quite a decent amount of material to have a decent cooling power.

 

[EatenByLimestone]

How does ZT figure into solar panels? Roofs get plenty hot. I can see an application where you want the outer surface of a roof hot, but the inside cool.

 

[stoveliker]

What matters is at what temperature ZT is this optimum; many "records" are achieved at temperatures that are not useful for optimal cooling. (e.g. a record ZT around 140 F is not useful to extract heat to make electricity when your heat source is at 250 F, as your chimney...).

I did not see a temperature mentioned in that Science news piece (tho I did not look back now and I might have missed it).

The point is ZT is dependent on the temperature.

Also, I'm not sure adding a thermoelectric would cool your panels much even if you extract electrical power - after all the convection (air cooling) at the back side now is nothing to sneeze at either. Covering the back side with an additional few layers might in fact heat the solar active device more, even if you are able to extract electricity from those additional layers.

 

[EatenByLimestone]

I was thinking of it more as a shingle material.

 

[stoveliker]

That'll be expensive, a tin selenide shingle. Not worth the investment for how much power you'd get out. And not likely durable up there.

 

[Corey]

The original article somewhat hid the actual scientific paper behind the "thermal conductivity" link, here is that and a few additional

Polycrystalline SnSe with a thermoelectric figure of merit greater than the single crystal - Nature Materials

SnSe has a very high thermoelectric figure of merit ZT, but uncommonly polycrystalline samples have higher lattice thermal conductivity than single crystals. Here, by controlling Sn reagent purity and removing SnOx impurities, a lower thermal conductivity is achieved, enabling ZT of 3.1 at 783 K.

www.nature.com

Tin selenide - Wikipedia

en.wikipedia.org

As far as selenium - yeah, not that great. But doesn't seem to be any worse than anything else. People don't seem to worry about InGaAs leds where the As is arsenic, a lot of the new 'promising' perovskite solar cells are based organic lead compounds. Even the current TEGs are based on tellurium, which is not exactly benign. Though I think the big key is that these are all compounds - so much the same way you might not want a cylinder of pure chlorine or highly reactive sodium in your house, you're perfectly happy luggng a 50 pound bag of water softener salt around - or even putting salt right on your food.

 

[stoveliker]

I agree (and this goes a bit too far into the weeds), but Se (more so than Te) has a habit to escape. GaAs is a covalent compound, strongly binding As. SnSe is much less "capturing" of Se.

 

[jetsam]

In general, papers implying big gains from novel materials or methods are a dime a dozen.

It is one thing to say "the predicted energy density of this chemistry is twenty times that of existing lithium batteries", and another thing to consider how these hypothetical batteries will perform in terms of cf discharge rate, cycle life, safety, and cost of manufacture at scale.

So it's okay to get a little excited about a novel material or method, but don't imagine that a new consumer product will be right around the corner as a result of the research.

 

[stoveliker]

Yep. The difference between basic research (the material needs to be found) and R&D that a company might fund in hopes of financial gain.

 

[HiroariTillWhen]

It's likely to take some time before the technology is commercially available and becomes interesting enough to buy, but when the day comes, it'd be a great material to line the walls of the place where your fireplace/boiler/furnace is located!