For this metal, electricity flows, but not the heat | Hacker News Reader

For this metal, electricity flows, but not the heat

sciencebulletin.org

123 pointsupen25 comments

Because of the material's uniquely diametric state transitions and properties, it gives itself naturally to negative feedback (self-regulating systems) -- for example, as a coating in space windows that reflects more radiation as the temperature increases. Not to mention the regulation of heat/electrical conductivity which is the primary subject of the article, in which the material behaves like an exotic semiconductor:

“This material could be used to help stabilize temperature,” said study co-lead author Fan Yang, a postdoctoral researcher at Berkeley Lab’s Molecular Foundry, a DOE Office of Science User Facility where some of the research was done. “By tuning its thermal conductivity, the material can efficiently and automatically dissipate heat in the hot summer because it will have high thermal conductivity, but prevent heat loss in the cold winter because of its low thermal conductivity at lower temperatures.”

I don't think this would work for coatings. The dimension and geometry of the material matter a lot and this is evidenced by the fact that they know that VO2 has normal behavior in polycrystalline thin-films. The VO2 they studied is single crystal and shaped into a nanobeam, meaning that already by geometry there's a huge shape anisotropy that I'm sure influences the properties. The beam is essentially suspended as well so it's relatively free of strain. Honestly I don't think this is going to leave the lab, you're not gonna find this behavior in bulk, only in nanoscale constructions.

Also the title of this submission is wrong, heat does flow, just at a lesser magnitude than expected from the Wiedemann-Franz law.

That... is the great breakthrough that many people were waiting for in thermoelectric generation.

Tell me it can be easily manufactured and we have a winner. This has the potential of making solid state thermal plants, where no moving parts would be required to transform heat into electricity.

Even if it's not particularly easy to manufacture it would still be massively beneficial to a lot of generator designs. We already do some crazy stuff to make single metallic crystal turbine fins for jets, I could see something similarly exotic in manufacturing being useful for power plants even if it costs 10x-50x more than the traditional parts given what this material can enable.

Would it be possible to use such a thermo-electric generator as a heat sink for a chip?

Depends what you mean. If you want to use it for power generation then you can but it is probably a bad idea: it gives resistance to the heat flow.

However, the thermoelectric effect is used by overclockers but on reverse : they consume energy to move heat from the CPU to a radiator. It is called a Peltier cell.

No, as you need to be removing heat from the hot side anyway.

How about its use in computer chips? Maybe all those fans and heat sinks won't be needed.

I think it's the other way around: the chips are producing heat constantly and you want that heat to dissipate as quickly as possible, so it's not the right place to put a thermal insulator.

It might interesting to note that diamonds work the other way round, fantastic insulators of electricity but good conductors of heat.

Ceramics too.

I used to have to work with transistors backed with BeO (beryllium oxide) ceramic. A very scary material, inhaling the dust once can cause a lung condition for life. It is second behind diamond in terms of electrical insulator, thermal conductor.

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

link: http://newscenter.lbl.gov/2017/01/26/electricity-not-heat-fl...

I wish they included some more numbers. The paper is paywalled: http://newscenter.lbl.gov/2017/01/26/electricity-not-heat-fl...

I think Fig 1 on page 3 applies: https://arxiv.org/pdf/1601.06246.pdf so it's 100ohms/cm vs 0.0001ohms/cm for the best example. Still looking for thermal conductivity.

Those electrical properties sure stood out as being superconductorish.

Could this be used in CPUs? And would the cost be affordable?

Why would you want a CPU that can't get rid of its heat?

Maybe I misunderstand.. but isn't it allowing electricity to flow without heating ?

No, the electricity flowing will still generate heat. Flipping bits may not take a lot of power, but it does take power, and some of that power is converted into heat regardless of what material you use. Except maybe superconductors, but I'm not sure.

To amplify a bit on what ranger207 said: Normally, metals are good conductors of heat, for the same reason that they are good conductors of electricity - the flow of electrons transports the heat.

In a computer chip, the heat generation isn't (primarily) due to the flow of electricity along the conductors - it's primarily in the transistors (and IIRC, mainly in their switching).

Given some of the comments here I think it's important to note that the thermal properties of Vanadium are well known and the metal has been used for years in high-speed steel and alloys, such as Chrome Moly Vanadium, where heat is the enemy.

This discovery in regard to conductivity could be game changing, but it's not really a surprise.

The way they describe the properties of it make it sound like a superconductor. Am I missing something?

This is great, I wish I had it on my windows now.

If this could be applied to greenhouses in cold climates that would he amazing. A glass/polycarbonate coating that lets in the Sun's radiation but restricts heat from leaving.

Current greenhouse insulators already do that, and very well. I don't think this material would be worth it.

But for electricity generation things are very different.

This is basically unobtanium [1]

[1]https://www.youtube.com/watch?v=b_HhiU1mOwU

Reply on news.ycombinator.com