Originally posted by humy
I don't pretend to really understand the physics of this so this might be a stupid question to someone who does but, I wonder; theoretically, can this somehow be used to in effect locally cool certain nano scale components, such as nano-scale infrared sensors or superconducting nanowires ...[text shortened]... ture even though the macroscopic environment that component is in is at around room temperature?
That would depend on how low a temperature you are contemplating. We use several cryopumps where the cold head is sitting very close to zero Kelvin, about 10 degrees Kelvin. It is cooling some activated carbon pellets about the size of rabbit feed. This carbon has an extremely large surface area and when a molecule from the chamber being pumped out hits the carbon, it checks in but does not check out, like the roach motel, except when it warms up in a process known as 'regen' all those molecules exit the pellets and have to be pumped out by a mechanical roughing pump.
The thing that keeps the pellets at 10 degrees K is it is in a vacuum.
You would never in a million years get anywhere near that temperature at atmosphere because the air around the cold part would suck away heat faster than it could be pumped out.
So IF you had the parts in a decent vacuum, like around 1E-6 millitorr, heat only transfers by radiation and radiation is a known poor transmitter of heat so you can get down to low kelvin numbers in such a vacuum.
There also is a vacuum guage, called a Thermocouple guage that has a short thin wire inside which has a TC gauge hooked to the center of the wire.
It works at a vacuum level much closer to atmosphere than the cryopump which can reach 1E-10 millitorr. My machines SOMETIMES gets to 1E-8 millitorr, 2 orders of magnitude worse vacuum level than a good system can achieve but it is way good enough to sputter ions off a target for our coating needs.
The vacuum levels a TC gauge works at is between 1 micron and 1000 micron (same as millitorr) and what happens there is at the lower end of its vacuum scale heat is only conducted away from the wire by radiation, IE, not very well, so the wire gets very hot, 200 to 300 degrees C. The thermocouple reads that temperature and it is very well calibrated so they know there is a better vacuum in it than the top level of 1000 Millitorr.
At 1000 millitorr, the temperature of the little wire is around room temperature because at that level, heat is taken away by direct conduction to the large amount of molecules surrounding the wire. So the thermocouple measures the temperature of the wire very accurately and the meter hooked to the electronics registers a higher current and goes to the top of the scale.
The scale is calibrated in millitorr, from say, 1 millitorr to 1000 millitorr and reads fairly accurately at that level of vacuum. But you see, as you get more and more molecules surrounding the wire the lower the temperature gets over what it COULD be in a very low vacuum.
So a part could be at room temperature on one end and 10 K on the other IF IT WAS IN VACUUM.