New Material Breaks World Record Turning Heat into Electricity (opens in new tab)

You'd be actively cooling your wood stove by much more than the power you'd extract, and you won't get more than enough to charge a phone or laptop. TECs are just terrible.

You'd get much better result by making a small steam plant with the same setup: boil water on stove, drive plant (turbine or piston), condense outside, repeat.

However, if you really want to do this with TECs, stack them to lower the per-unit temperature differential, or distribute the heat energy over a larger area and run them in parallel.

There are some single layer units that can get up to ~90 of delta_T, but I don't think they're worth dealing with. As a sibling comment points out, you can stack TECs - indeed, you can buy integrated units with very high delta_T that are just pre-stacked TECs.

I think it'd certainly be fun to play with. Some bits of warning - most TECs are physically quite fragile. Not in the sense that you need to be super careful when handling them with your hands, but that heat stresses can totally wreck them. Do NOT hard mount both sides (hot and cold) to different materials - the differential thermal expansion on the two sides has a potentially of ripping your TEC apart. You may be able to get away with a hard mount (such as solder) on one surface, and some sort of thermal transfer material on the other.

Also warning, your hot side is hot enough to start doing weird things with normal solders, which might be the bigger limiting factor.

Won't this reduce the thermal output of your stove quite significantly? I suppose that might not be a problem in practice (most wood stoves I've used end up overheating the room if you're not careful), but it's not like it's free energy :P

Hello fellow Canadian :)

Wouldn't a home made Sterling Engine be easier and cheaper? Then use the motion to power a small generator.

But like others commented you may just end up cooling your stove. The Yukon can get pretty cool.

Maybe an old washer motor, a prop to generate power from wind? Or solar in the summer when light is ample.

That's a cool idea.

I came across this link because I want to build cheap small sensor stations using LoRa to transmit data (gps location?, temp, humidity, air quality) to a central server.

Solar is not really an option because it needs to be cleaned once in a while.

Tear one of these apart? https://www.amazon.com/dp/B007BLN652/ref=cm_sw_r_cp_api_i_5j...

You could also use the coolant to pull heat off the stove (obvs with appropriate safety built in) and run the hot side of the TEC within spec.

I’ve seen wood stoves get red hot after an incorrect damper setting runs for 5-10 mins, direct heating might fry your circuits.

I've been experimenting with this idea too. The idea is to drive a fan, which blows on metal heatsink attached on the cool side of the TEC. The air stream further accelerates the fire, so this becomes a reinforcing circle. Directly wiring it is not enough, you need a battery to start, buffer, then you need a controllers, never had time for this

There is a company making beautiful camping stoves with TECs and fans: www.bioliteenergy.com

How much power are you trying to make? You can already buy "off the shelf" USB power supplies for camp fires

Co-generate with a Stirling engine or steam turbine. Your application is the wrong scope for TECs. It will be a hundred times more expensive, and a tenth of the efficiency.

Could you chain 3 of them using a tank of coolant at room temp as the between value. Or make it water and put the TEC on the cold side of the tank

Solar & lithium batteries are so cheap from scale, it's worth comparing the cost, even though production is very low in winter.

I am twenty years removed from it, but I used to be well-informed on this kind of thing; I have a PhD in mineral physics.

Let's see how well I can explain this (haven't read the article, yet, sorry! Waiting for a plane...)

So you're no doubt familiar with the physics of a vibrating string; it resonates at wavelengths (length of string, 2 * length of string, 3 * length of string... n as n->inf). So you can express any vibrational state of the string as sum(intensity * wavelength); so you can represent the state of the string as a vector of intensities on a basis of allowable vibrations of the string.

Let's call these _vibrational modes_. Let's assume occupancy of these modes is quantized. It's (sort of...) the same as energy levels of atomic orbitals in electronic structure, if you remember that from chemistry classes; the way it's not the same is important (bosons vs fermions) but not at this level of handwaving :-)

So this is how solids store energy, and we call this energy "heat".

A reasonable approximation for a crystalline structure is balls – point masses – connected by springs, where the springs are covalent bonds, plus electrostatic effects between point charges. Intuitively, you can follow that the same kind of _vibrational spectrum_ will arise from this arrangement (in the same kind of way; the solutions of the differential equation of this system of forces under periodic boundary conditions). So materials have resonant frequencies in the same way guitar strings do.

Therefore, this vibrational spectrum defines the thermal behavior of a material; heat capacity, thermal conductance, etc etc etc etc. Each of these vibrational modes is also tied to a collective motion of the particles in the material, which (if sufficiently violent) will tear the structure apart – there's the solid-to-liquid phase transition – or, more subtly, if lost will lower the symmetry of the crystal structure, which gives rise to solid-to-solid phase transitions (an example would be alpha to beta quartz, which will crack your crockery if you leave it in the oven on a cleaning cycle; https://en.wikipedia.org/wiki/Quartz_inversion).

There's a lot of depth here, as you can imagine!

The space of "fairly typical materials" is very wide because there are lots of elements and their combinations blow up very fast. Secondly, materials are often very sensitive to small physical or chemical changes, resulting in wildly differing properties.

To you the simpleness of the final result is surprising, but that simple result was discovered after a long and exhaustive search. A search into a wide and shallow space can be just as impressive and difficult as a search into a narrow and deep space.

> a recipe of fairly typical materials

We don’t have good theory for materials science. Our understanding is closer to a list of observations (“effects”) than anything unifying. That almost ensures there will be surprises in the gaps.

At the same time, popular disillusionment with repeated claims of wonder materials has led to—in my opinion—underinvestment in basic research.

But there are some pretty cool applications. Apparently the heat difference between a buried water-pipe and the surrounding earth is already enough to power a wireless sensor, which can transmit data to localize leaks, for example.

From mucking about with the device efficiency formula from wiki (https://en.wikipedia.org/wiki/Thermoelectric_materials#Therm...), and a change of zT from 2.5 to 5, we see a maximal possible efficiency increase of 38% (when T_c == T_h).

Wiki also tells me that the best TEG modules currently lock in around 8%, so we're looking at like 10-11% at best with the new material.

So from a bulk scale electricity standpoint... the needle probably hasn't shifted at all. From a small scale? In the IoT like applications (as mentioned in the press release), that extra 30-40% is nothing to sneeze at.

The Peltier and Seebeck effects are so grossly inefficient that even an order of magnitude increase in efficiency doesn’t bring it into reason for basically any purpose. I’m not actually aware of any device ever made that uses the seebeck effect in any substantial way other than the little heat powered fans people put on wood stoves.

The peltier effect is just down right awesome, you put power in and now it’s cold!? Reality steps in at some point when you need to drive down the efficiency even further to get large differentials and ugh. They’re insane to deal with, any amount of thermal load worth speaking of means you have to use a phase change system.

It is very handy for camera sensors though and other scientific gear. There’s a world world of CCD sensors that act in a vacuum with peltier devices driving them below -30c to reduce the noise produced by the sensor.

it works similar to a thermocouple. A normal thermocouple needs a voltage bridge to increase the voltage with a millions to get anything that can be read by a digital 5v meter. These materials don’t need that. And a couple of degrees could produce a workable voltage. But the title Of the article states that the material is meta stable, unsure if that means it will work for a couple of seconds then die. Also, vanadium is used which makes it quite costly.

The predominant use cases for this will be powering ultra-low power sensors. The higher power could enable better (longer range, higher bandwidth) radios in the sensors.

I think you are thinking about an RTG instead of an actual reactor.

Unlikely. Much easier to just connect the sensor to the main power supply via a wire.

Yes, the thermal-electric effect is reversible. It's sometimes called the Peltier effect (or thermal-electric cooling) when used in heat pumping mode, and Seebeck effect when used in electricity generation (thermal-electric generation).

If you have a very small cooling job, yes. Anything above a handful of thermal watts is still better off using a compressor.

Yes, the thermoelectric effect works both ways.

> Would they be able to capture enough energy so their star would be indistinguishable from the ambient temperature of space?

You would still see infrared:

https://en.wikipedia.org/wiki/Fermi_paradox#Conjectures_abou...

> Such a feat of astroengineering would drastically alter the observed spectrum of the star involved, changing it at least partly from the normal emission lines of a natural stellar atmosphere to those of black-body radiation, probably with a peak in the infrared.

https://en.wikipedia.org/wiki/Technosignature

> A Dyson sphere, constructed by life forms dwelling in proximity to a Sun-like star, would cause an increase in the amount of infrared radiation in the star system's emitted spectrum. Hence, Freeman Dyson selected the title "Search for Artificial Stellar Sources of Infrared Radiation" for his 1960 paper on the subject.[4] SETI has adopted these assumptions in its search, looking for such "infrared heavy" spectra from solar analogs. From 2005, Fermilab has conducted an ongoing survey for such spectra, analyzing data from the Infrared Astronomical Satellite.[5][6]

If we look at our planet, we should realize that fossils and nuclear power are limited. Aside of the quantitative limitation, we get to feel the negative green-house effects now, making alternative energy sources even more urgent. The only real long-term source is the sun. We HAVE to (completely) switch to sun energy within the next 100 years. This will require huge photon harvesting facilities. For now, covering our deserts would be enough. But what if our energy requirement increases 10 fold? The more energy we have, the more people can have steaks and luxury. In the SECOND we reach the point of energy limitation on Earth, we will expand our photon harvesting facilities into space, close to Earth. This will grow with generations, so will technology. The gigantic amount of energy daily shining onto Earth is still ridicously few compared to what the Sun radiates every second. The photon harvesting layer will shrink, become more efficient, harvest more wavelengths (maybe even just a single atomic layer) and routed to central cores where it is collected and sent to Earth or the future intelligence habitat. Even if it is just a mirror, redirecting the photons to a central core, which routes it close to Earth, where it will evaporate water - or whatever much smarter ideas our future generations will find.

And this is what we could do now with our technology.

The infra-red issue, imho, is based on current technology level. No light enters the depth of the ocean because life evolved to build photon-recepters of all wave-lengths, even the less fruitful.

Taking this into consideration, every maturing civilization WILL inevitably produce Dyson spheres at least transiently. We already started it! Just look at the solar moduls of the ISS, which are actually the beginning of a Dyson sphere.

What would they use all that energy for, and how would they use it in a way that wouldn't radiate heat or light on par with the amount captured?

Sounds like a good beginning to a collaboration novel by Neal Stephenson and Alastair Reynolds.

You can use these types of materials to extract energy from any energy gradient.

In the particular case of solar panels, in order to have a workable thermal gradient, you need to have some sort of conductive path from the solar panel to the cold reservoir (assume the ground under it) - that's extra cost there. Then once you have the TEG running, what you're actually doing is adding an impediment to heat flowing from the solar panel to the cold spot, likely causing the solar panel to be a bit warmer (thus decreasing its efficiency). The increase in energy production per given investment is almost always going to be lower than just getting more solar panels.

You see this is most bulk energy production contexts. It's rare for these energy scavenging techniques to make economic sense. Where you start seeing them make sense is when you have other constraints come into play. You can see this with other aspects of solar generation like solar tracking.

The title is misleading, the technology (thermoelectric cooler) does not turn heat into electricity, but temperature gradients.

It needs a nearby source of cool to work, and when a solar panel is hot, oftentimes everything surrounding it is hot.

Commercial vanadium metal, of about 95% purity, costs about $20/lb. This is a tiny layer, so even at the 99.9% purity level cost of $100/oz, materials cost is not likely be a critical factor. The main constraint is that it's not making much power -- at this level it's really for very low power applications, especially things like sensors and comms.

I don't think the research is intended for large scale electricity production. It's for small sensors and processors and that won't cost much.

This opens up a quiet a few possibilities like buy a temperature sensor place it somewhere and it starts feeding data into your home WiFi. No wires, no batteries.

Easy, just build a fusion reactor capable of eventually producing vanadium from hydrogen, and you got yourself a stew going!

Edit : Method of producing said reactor left as an exercise for the reader.

Vanadium is only $6 per pound. About as expensive as street price of ice cream. It's not a rare earth element or anything like that.

I'm glad you ran the numbers for this to reach such an important conclusion about the future of this technology, RandomWorker. Care to show your working?

No you can't make a perpetual motion machine with it ;)

Are you asking if it violates the 2nd law? Of course not.

It's just a heat engine like any other electrical generator so it's not going to have fundamentally new kinds of applications.

Don't forget you can also use the waste heat from a device (like a computer) to run a mechanical heat engine to generate power to help run the device. But only partially. I guess we don't do that much because it's probably not economically feasible. That's kind of what turbochargers in cars do though.

This will be slowly but surely be used everywhere. Reducing heat output while providing electricity is almost always a win-win situation.

I think it's correct. Heat is the flow of thermal energy. A portion of that flowing energy is converted to electrical energy. Perhaps you're confusing heat with thermal energy?

There must be some heat flowing to generate electrical power because of the 1st law, so even if the effect is described as being due to a temperature difference, in practice, you also need a heat flow to be useful.

Your quote about "without any side effect" means without heating up a cold reservoir. But the article doesn't claim that.