Unexpected nanotube heat transfer suggests new way to cool processors

A strange phenomenon observed at the University of Maryland could pave the way to new techniques for cooling electronics: when researchers passed a current through carbon nanotubes, they didn’t heat up – but other nearby objects did. While still studying exactly what mechanism produces the phenomenon, the researchers report …

[Anonymous Coward]

Heat sink you say....or heat bridge?

This as a heatsink would be very very nice indeed, though it is in effect a passive peltier and just shows that the magic carbon nanotubes as well as conducting electricity, also conduct heat very effeciently. So effeciently that they end up heating the surrounds and what is supprising is that they will pass that heat on to the extent that the surrounds will become hotter than the nano-tubes and it is this that does not make sence with the current version of science events.

One way to think of them is like a water pipe, that when turned off empties all it's water into the surrounds. Another way is to think of it as heat having it's own form of polarity and that carbon nanotubes are more neutral than other matriels so the heat gets attracted to them as apposed to all other materials that have the same polarity/charge and as such all things become equal.

Now whilst this may raise a lot of questions and future diplomer course's I have one real simple, answearable one. That is the question of feedback and in that I mean is the heat once the heat source is turned off balancing all around or is the flow of heat directional. This is a important question as it will effect things like it being used as a heatsink. If it is as I suspect by the design of the tubes omni-directional, then the need to passivily cool the large block metal the nanotubes are disapating in after the system has stopped producing heat for a fair while. Reason being that the heat in the disapation heatsink will feedback into the nanotubes and back onto the CPU. Thus potentualy overheating the CPU and damaging it after the system is shutdown. This is based upon them being heatpumps and needing somewere to pump that heat and disapate it greater than the amount of heat being generated by the CPU. Now if there directional - happier day's and I look forward to the day when they vent plasma :-).

Also if these nanotubes are non-directional then perhaps they could try carbon nano-cones. Also what are carbon-triangles like compared to nano-tubes. Could have tube that tapers to a larger square end; That I'm sure would induce some form of directional heat transfer and if it doesn't then it will only fill in some of those many blanks about this old-new field of science.

But good news, a material that conducts electricity has been found to also conduct heat and as it was realy effecient at conducting electricity, it also conducted heat realy effeceintly beyond what was expected.

[G C M Roberts]

Might I be the first to say I for one hope similar research has been done for Copper Nanotubes. In fact I will do a googlefor hot CuNT right now to check.

Mine's the coat with an awkward bulge.

[topcav]

So what happens when you pass a current through nano tubes which are just suspended in a vacuum? There would be nothing to be heated by any generated fields so do they heat up themselves or is there simply no loss?

[Whitter]

In a vacuum.

You'd need to provide a means of doing so first, which likely then provides the alternate heat sink/source material(s).

[Charles 9]

How about outer space?

Insignificant gravity and no atmosphere. What would happen if you did the trick with no substrate in space?

[John Smith 19]

Possibly the *best* kind of science.

Go looking for something.

Find something else.

Realize something weird is going on.

Intriguing. No idea where it will go but sounds like it could have some legs.

Thumbs up for good science

[Anonymous Coward]

Anyone know if the effect is reversible? As in Heat In: Current Out ?

[Robert Carnegie]

Reversible?

I don't think that's physically possible, but that's why you asked. In fact I don't think this is about transferring heat at all, but controlling where heat appears. In that case, for a microprocessor, it would be not so much a heat sink, as a structure where processing occurs inside the core of the component but heat is generated on the surface. I'm visualising a 1970s fibre optic lamp made of nanotubes, which is probably wrong. But pretty.

[Ru]

We can already do that

We've plenty of handy things for converting heat into electricity... thermocouples, peltiers, solid state and mechanical sterling engines, plain old heat exchangers and turbines and so on.

But in this case... no. The nanotubes are neither generating significant amounts of heat, nor magically absorbing it. They're 'simply' acting as a thermal bridge.

[ItsNotMe]

“remote Joule heating”

This have anything to do with online porn?

Mine's the tan rain coat in the corner...thanks.