Absolute Zero

[Minerva]

If ALL motion stops at absolute zero is it just theoretical or just an absolute value upon which a condition can approach even in the distant regions of the universe?

 

[f95toli]

All motion can't stop since that would violate Heisenbergs uncertainty relation, you still have what is known as zero-point fluctuations (which can be measured).

One problem when talking about this is that "temperature" is not a very well defined concept when you start to approach 0 K. When can cool down gases to nanokelvins using laser cooling but in reality we are then just using temperature as a measure of a particle's kinetic energy (how much it "vibrates"); you can just as easily use joules or even frequency (depending on what you are doing).

Even at more "normal" temperatures like a a few tens of milikelvin (which can be easily reached using a dilution refrigerator) temperature is not very well defined since i.e. the electrons can be at much higher temperature than the crystal lattice (due to the reduced electron-phonon coupling).

So the answer to your question is: It depends on what you mean.
We can be quite sure that the kinds of temperatures we can reach in the lab could never occur naturally, there is way to much energy around in the natural world.

 

[Minerva]

What about in space far away from any source of energy?

 

[f95toli]

The problem is that even very faint light from remote stars is enough to heat particles. Moreover, there is still some the remaing radiation from Big Bang and that radiation is everywhere and,again, will heat.

 

[Minerva]

Yes that's what I'm seeing.

It would be amazing if it could be created artificially. A full stop. :Q

 

[f95toli]

As I wrote above: You can never "stop" motion complettely, it is literally impossible. There are ALWAYS zero-point fluctuations. There are plenty of experiments where zero-point fluctuations are important so the theories are well tested.

 

[SuperFungus]

I always thought that even outside Heisenbergs uncertainty relations 0 K is impossible, naturaly or artificially, because there is no perfect insulator to contain something in; some small amount of energy will always find its way through. In fact to get to 0 k insulators would be useless, unless they too where at 0 k. I'm not even aware of a way to measure the temperature without somehow injecting a little energy into the object you're trying to cool. So I agree with f95toli completely, not only is 0k impractical but it is impossible.

 

[CycloWizard]

I don't recall exactly how 'absolute zero' temperature was determined, but if I were to design an experiment to measure it now, I would try to approach it asymptotically using the methods f95toli suggested. Even if you couldn't actually achieve it exactly, you could still tell where it occurred if you collected your data appropriately.

 

[DrPizza]

IIRC, a simple experiment to estimate absolute zero had to do with the ideal gas laws. If I remember correctly, you compare the volume of a container of gas when immersed in boiling water and when immersed in ice water. Graph the volume vs. temperature, draw a line, go backwards to where v = 0, and the temperature is pretty close to absolute zero. I've also seen an experiment similar to this done using the pressure.

edit: I think that in the past, since there was plenty of experimental evidence to see that there was a direct relationship between temperature and volume, it was first inferred that since you couldn't have a volume <0, there was an absolute lowest temperature. I'm not certain about this point though.

 

[f95toli]

Cyclowizard: Temperature is a statistical property, in thermometry it is generally just "defined" as part of the exponent in the distribution function, i.e. exp(-U/kT) where k s Boltzmanns constant (at low temperatures the Fermi- or Bose-Einstein distributions must be used but the same exponent appears there).
This works even when measuring the temperature using electronic sensors (i.e. silicon diodes) since these Fermi distribution determine the number of excitations in the conductions band and therefore the resistance of the diode.
In a fully quantum mechanical treatment of a system "temperature" often enters as a "thermal bath", an ensemble of harmonic oscillators with a disitribution of energy that follows the Bose-Einstein distribution at a temperature T.

As I have already pointed out, in the case of low temperatures (e.g. laser cooling) the concept of temperature becomes ill-defined but can still be used as a measure of the kinetic energy or more generally the amount of noise in the system, the latter is not always directly related to the thermal noise. I.e. the temperature measured by an wide bandwith electronic sensor is quite often the thermal temperature plus the electrical noise in system. This is a VERY important point which is often overlooked even by many scientists, I have seen examples of experiments where they have claimed that the temperature was e.g. 50 mK (the temperature of the fridge) but the where the real ("relevant") temperature was in reality much higher, in some cases several Kelvin which has a significant effect on the experiments.

Also, it is interesting to note that "noise temperature" is often used as a measure of the the noise properties of e.g. amplifiers and quite often this is LOWER than the thermal noise, you can e.g. buy microwave amlifiers that have noise temperature of 4K when operated at a temperature of 15K. The reason why this is possible if of course that they have a limited bandwidth (thermal noise is "white", meaning the bandwith is infinite).

Hence, the word "temperature" can be used in many different ways and it is only at relatively hight temperatures that these uses coincide so some extent.

 

[RSMemphis]

The rate at which you can extract heat from an object approaches zero as you reach the absolute zero point. Therefore, it is impossible to reach absolute zero, but it is possible to approach it as closely as desired (with a lot of energy).

 

[silverpig]

Actually, the more sensical quantity in thermodynamics is inverse temperature 1/T.

1/T = dS/dE

Inverse temperature is the rate of change of entropy with respect to energy. So, when you add energy to a thermodynamic system, how much does the entropy change? Of course you can just look at dE/dS, but then you are asking how fast does the energy of a system change as you increase its entropy. It's a little harder to think about that way.

It also makes more sense to use 1/T to describe temperature as absolute zero temperature would be infinite inverse temperature, which make more sense why we can't reach it.

What's also interesting is it is possible to have something at a temperature below absolute zero in certain contrived cases. What happens is the system then cools to negative infinity, flips over to infinity, and comes down from very hot to room temperature. This is possible as you just can't have something AT absolute zero, but you can have systems below absolute zero

 

[f95toli]

Silverpig: Yes, but the problem here is that classical thermodynamics doesn't work very well for real systems when the temperature approaches zero since phenomena like Bose-Einstein condensation etc becomes important; a full quantum mechanical description is therefore needed and classical thermodynamics breaks down.

Trivia: The international temperature scale (ITS-90) is only defined above 0.7K, hence strictly speaking "temperature" is not defined for temperatures lower than that.

 

[silverpig]

Originally posted by: f95toli
Silverpig: Yes, but the problem here is that classical thermodynamics doesn't work very well for real systems when the temperature approaches zero since phenomena like Bose-Einstein condensation etc becomes important; a full quantum mechanical description is therefore needed and classical thermodynamics breaks down.

Trivia: The international temperature scale (ITS-90) is only defined above 0.7K, hence strictly speaking "temperature" is not defined for temperatures lower than that.

For sure. Going to large inverse temperatures is sort of like going to high velocities in Newtonian mechanics... things start to depart from the classical answer.


BTW: What are you working on again? IIRC it was lasers into low pressure sodium gas for some kind of quantum coherence experiment right?

 

[CycloWizard]

Originally posted by: f95toli
Cyclowizard: Temperature is a statistical property, in thermometry it is generally just "defined" as part of the exponent in the distribution function, i.e. exp(-U/kT) where k s Boltzmanns constant (at low temperatures the Fermi- or Bose-Einstein distributions must be used but the same exponent appears there).
This works even when measuring the temperature using electronic sensors (i.e. silicon diodes) since these Fermi distribution determine the number of excitations in the conductions band and therefore the resistance of the diode.
In a fully quantum mechanical treatment of a system "temperature" often enters as a "thermal bath", an ensemble of harmonic oscillators with a disitribution of energy that follows the Bose-Einstein distribution at a temperature T.

As I have already pointed out, in the case of low temperatures (e.g. laser cooling) the concept of temperature becomes ill-defined but can still be used as a measure of the kinetic energy or more generally the amount of noise in the system, the latter is not always directly related to the thermal noise. I.e. the temperature measured by an wide bandwith electronic sensor is quite often the thermal temperature plus the electrical noise in system. This is a VERY important point which is often overlooked even by many scientists, I have seen examples of experiments where they have claimed that the temperature was e.g. 50 mK (the temperature of the fridge) but the where the real ("relevant") temperature was in reality much higher, in some cases several Kelvin which has a significant effect on the experiments.

Also, it is interesting to note that "noise temperature" is often used as a measure of the the noise properties of e.g. amplifiers and quite often this is LOWER than the thermal noise, you can e.g. buy microwave amlifiers that have noise temperature of 4K when operated at a temperature of 15K. The reason why this is possible if of course that they have a limited bandwidth (thermal noise is "white", meaning the bandwith is infinite).

Hence, the word "temperature" can be used in many different ways and it is only at relatively hight temperatures that these uses coincide so some extent.

Good to know. I really know very little (read: just about nothing) about very low 'temperature' systems, as I tend to dwell in the 37°C range with everything I'm doing right now. My background in quantum anything is limited to a single course taken about 5 years ago.

 

[superHARD]

Stupid person injection

Isn't there lots we don't know for sure?...what if black holes suck everything from close by objects?...even sucking away all heat?

Don't belittle me too much please...

 

[gsellis]

Assuming Big Bang, I say about 30billion light years past the edge of the universe might be absolute zero. It would be completely void of any matter or energy. In theory of course.

 

[andrewbabcock]

Originally posted by: gsellis
Assuming Big Bang, I say about 30billion light years past the edge of the universe might be absolute zero. It would be completely void of any matter or energy. In theory of course.

The edge of the universe? There isn't one...The universe is infinitely large...

 

[gsellis]

Originally posted by: andrewbabcock

Originally posted by: gsellis
Assuming Big Bang, I say about 30billion light years past the edge of the universe might be absolute zero. It would be completely void of any matter or energy. In theory of course.

The edge of the universe? There isn't one...The universe is infinitely large...

errrr... I meant the expanding part from the bang. So yeppers

 

[superHARD]

Originally posted by: andrewbabcock

Originally posted by: gsellis
Assuming Big Bang, I say about 30billion light years past the edge of the universe might be absolute zero. It would be completely void of any matter or energy. In theory of course.

The edge of the universe? There isn't one...The universe is infinitely large...

Isn't that theory up for grabs?

 

[bsobel]

Originally posted by: gsellis
I say about 30billion light years past the edge of the universe

I say you flunked physics

 

[bsobel]

Originally posted by: superHARD
Stupid person injection

Isn't there lots we don't know for sure?...what if black holes suck everything from close by objects?...even sucking away all heat?

Don't belittle me too much please...

Black holes generate alot of radiation, they are a source of heat, they do not remove it.

 

[beansbaxter]

All motion in the universe or just on this planet?

 

[Biftheunderstudy]

Absolute zero would be local so the universe wouldn't freeze. The average temperature of the entire universe is pretty easily measured from the Cosmic Microwave Background Radiation ~2.7 Kelvin. There are lots of interesting quantum mechanical things you can do with temperature, like negative temperatures. Keep in mind that negative temperatures are "hotter" than positive temperatures. Going below absolute zero is not defined since theres an infinity there.

BTW f95toli whats your area of research, I was a research assistant over the summer working with laser cooling of sodium.

 

[superHARD]

Originally posted by: bsobel

Originally posted by: superHARD
Stupid person injection

Isn't there lots we don't know for sure?...what if black holes suck everything from close by objects?...even sucking away all heat?

Don't belittle me too much please...

Black holes generate alot of radiation, they are a source of heat, they do not remove it.

Has that changed? In my college physics class (3 years ago) I think that was just a "theory"