Why is absolute zero considered to be asymptotical? Wouldn't regions such as massive gaps between galaxy clusters have temperatures of absolute zero?

I just do not see why our model must work the way that it does. I mean there have to be regions with no thermal energy out there, the universe is massive.

  • 3
    $\begingroup$ Out of curiosity, can you imagine a place in the universe so remote even an infinitely powerful telescope cannot see any stars or galaxies? Because light carries energy. $\endgroup$ – user10851 May 29 '14 at 0:16
  • 1
    $\begingroup$ @ChrisWhite Thanks to accelerating expansion, I can. I'll show it to you, but you'll have to wait a few gigayears for me to set everything up. $\endgroup$ – rob Jun 12 '14 at 23:13
  • $\begingroup$ Related: physics.stackexchange.com/a/130381/51994 $\endgroup$ – alemi Aug 13 '14 at 0:41

We can only approach absolute zero asymptotically because we can't suck heat out of a system. The only way we can get heat out is to place our system in contact with something cooler and let the heat flow from hot to cold as it usually does. Since there is nothing colder than absolute zero, we can never get all the heat to flow out of a system.

We can reduce the temperature by increasing the size of the system and diluting the heat. In fact this is why the CMB (cosmic microwave background) temperature is only 2.7K rather than gazillions of K as it was shorlty after the Big Bang. The expansion of the universe has diluted the heat left over from the Big Bang and reduced the temperature. However achieving absolute zero this way would require infinite dilution and therefore infinite time, which is why the universe approaches absolute zero asymptotically.

Actually, assuming the dark energy doesn't go away the universe will never cool to absolute zero even given infinite time. This is because the accelerated expansion caused by dark energy creates a cosmological horizon, and this produce Hawking radiation. The Hawking radiation will keep the temperature above absolute zero.

  • 5
    $\begingroup$ Sorry nitpicking , but the the black body temperature at the time of separation fof CMB was about 3000k , less than what the sun black body is now.damtp.cam.ac.uk/research/gr/public/images/bb_history.gif $\endgroup$ – anna v May 28 '14 at 15:15
  • 32
    $\begingroup$ Astonishingly one gazillion is exactly 3,000K :-) $\endgroup$ – John Rennie May 28 '14 at 15:53
  • 6
    $\begingroup$ You mentioned that "the only way we can get heat out is to place our system in contact with something cooler" - but what about techniques that don't use thermal conduction, e.g. laser cooling? $\endgroup$ – voithos May 28 '14 at 17:45
  • 1
    $\begingroup$ I have also seen negative absolute temperatures appear when modeling systems that are not really measurable as thermodynamic temperatures, but the math happens to work out conveniently enough that the "negative absolute temperature" is used to convey the idea succinctly. $\endgroup$ – Cort Ammon May 29 '14 at 0:34
  • 3
    $\begingroup$ Dark energy with $w<-1$ would allow one to achieve infinite dilution in finite time. Although the side effect of spontaneous existence failure for the universe makes that a cold comfort. $\endgroup$ – Stan Liou May 29 '14 at 2:04

Current physics has established that the underlying framework is quantum. A basic principle of quantum mechanics is the Heisenberg Uncertainty Principle, HUP.

This ensures that for particles there can be no exact measurement or condition of 0 kinetic energy, so an ensemble of particles cannot have definite temperature of 0.

Photons are also particles and permeate the observable universe, even the emptiest of space, this is known as the left over radiation from the time that the universe became transparent to the electromagnetic waves . It is called cosmic microwave background, CMB, and its average energy is 2.7K.

This is what we observe:


All-sky map of the CMB, created from 9 years of WMAP data

There is no zero, although there exists a cold spot:

cold spot

The "cold spot" is approximately 70 µK colder than the average CMB temperature (approximately 2.7 K), whereas the root mean square of typical temperature variations is only 18 µK.

Thus, even if our universe is huge , there does not exist a spot without photons , and the HUP will ensure that there exists no zero K temperature.

Please also to remember that quantization of gravity will fill the space with gravitons, which will also be elementary particles and carry some energy, no matter how small, and will obey the HUP too. So in any interesting Universe there can be no 0K .

  • $\begingroup$ "there does not exist a spot without photons". Wouldn't this require an infinite amount of photons in any finite region of space? Or am I making a mistake in assuming photons are nice classical particles? $\endgroup$ – Kevin May 28 '14 at 16:10
  • $\begingroup$ @Kevin Yes, you are making that mistake. The number of photons is proportional to the energy in the region from E=hnu . Only infinite energy will give an infinity photons. $\endgroup$ – anna v May 28 '14 at 16:28
  • $\begingroup$ What about the interior of materials below their superconducting transition temperature? Wouldn't they be free of photons in regions sufficiently far from the surface? $\endgroup$ – Nicolau Saker Neto May 28 '14 at 16:50
  • $\begingroup$ @NicolauSakerNeto superconductivity is a quantum mechanical emergent phenomenon of the electromagnetic fields, i.e.photons $\endgroup$ – anna v May 28 '14 at 17:04
  • $\begingroup$ anna as much as i like your answer, alluding to gravitons (which like some assumed super-symmetric partners) might turn out to be completely different, adds an explanation gap $\endgroup$ – Nikos M. May 30 '14 at 22:40

Classical physics (intuitive) answer: temperature is a property of matter, that is a sort of estimate of the kinetic energy of particles. If you have no particles, there's nothing to measure the temperature of.

Mathematical answer: since temperature is changed by multiplying by a real number, the only way to reach 0 would be to have an object already at zero to begin with which would suck all the surrounding energy. This object has not been found to my knowledge (apart from gravitational singularities, and there are doubts about their existence).

  • $\begingroup$ And, by absorbing heat from its surroundings (or the thing you were trying to cool), the hypothetical object at absolute zero would cease to be at zero $\endgroup$ – David Richerby May 28 '14 at 19:56
  • $\begingroup$ Not even that, because every number multiplied by zero is zero :) $\endgroup$ – Bakaburg May 29 '14 at 14:33

In thermodynamics, temperature is defined from the zeroth law (yes, they thought of this after the first, second and third... - then realized they had forgotten something pretty fundamental). The zeroth law is this:

If system A is in thermal equilibrium with system B, and system B is in thermal equilibrium with system C, then system A is in thermal equilibrium with system C.

This is a fancy way of saying "temperature exists". But it also says "temperature exists when something can be in equilibrium with something else".

Thermodynamics really only deals with large scale response of a system - when you go down to the individual particle level you look at the kinetics of particles, and make statements about the "temperature" based on the average kinetic energy.

No region of space is "forever empty". While there might exist at any one moment in time a large region of space without particles in it, you can't actually know that it is empty - and the moment you discover it is not empty, you will have encountered a particle. If that particle was truly stationary (in which case - how did you come across it?), the act of measuring it will have imparted some momentum (uncertainty principle) so it's not at rest any more.

In other words - macroscopically you can't tell that an empty region of space has "zero" temperature; and microscopically, any attempt to measure zero will make it "not so".

  • $\begingroup$ this did not explain why it is NOT zero when not measured? and if this is so, does not answer why absolute zero Kelvin is only asymptotical $\endgroup$ – Nikos M. May 30 '14 at 22:37

The Universe is pervaded by the cosmic microwave background radiation, which has a temperature of $2.7$ K. Any region in the Universe must have at least this temperature.

  • 7
    $\begingroup$ What about the regions of the universe where we have achieved temperatures lower than 2.7k? $\endgroup$ – Michael May 28 '14 at 15:34
  • 3
    $\begingroup$ The Boomerang nebula is often cited as the coldest place in outer space we know and its temperature is as low as 1 K. The CMB radiation can only heat an object up to 2.7 K if the CMB photons manage to reach the object and be absorbed (and be given time to enter thermal equilibrium), which isn't always a guarantee. $\endgroup$ – Nicolau Saker Neto May 28 '14 at 16:10
  • $\begingroup$ So we would know we've left this universe if we find regions of space colder than 2.7 K? $\endgroup$ – TylerH May 28 '14 at 17:17
  • 1
    $\begingroup$ @TylerH, even if you rule out phenomena like the expanding gas of the Boomerang Nebula, you can't use the temperature to prove you're not in this universe. If you step through a magic portal and it's 1.5 K on the other side, maybe you're in a new universe, or maybe you're in our universe, a long time in the future where the CMB has cooled down. $\endgroup$ – Kevin May 28 '14 at 17:49
  • $\begingroup$ Or maybe you'll just wish you packed your thermal underwear. $\endgroup$ – Dawood says reinstate Monica May 29 '14 at 11:59

An empty space wouldn't have any particles to "be hot", but there would still be EM radiation of various wavelengths. Since no place can be infinitely far from any source of EMR, the radiation level can never drop all the way to 0. The CMB is everywhere, anyway.

Now, if there is no matter around, what are you taking the temperature of? You could just as easily claim that the temperature is infinite as it is 0 K. You need to carefully define just what it is that you're looking at the "temperature" of.


There has been many answers to that question and my answer will be probably lost in the sea of responses and different opinions. However, as it departs significantly from the proposed view points I might sketch it here.

Recent debates about heat flows, negative temperatures and the notion of entropy in statistical thermodynamics have allowed to raise questions about the notion of temperature. In particular, the negative temperatures (which although still debatable are quite well defined) seem to contradict the fact that zero temperature cannot be achieved. It also contradicts the intuition since negative temperatures are hotter than infinite temperatures which kind of does not make sense when said like that.

It turns out that none of these conceptual inconsistencies occur if instead of looking at temperatures, we look at inverse temperatures often denoted $\beta = 1/(k_B T)$.

In this case, heat always flows from low $\beta$ values to higher $\beta$ values regardless of their sign. Also the concept of absolute zero corresponds to an infinite value of $\beta$ and since there is no end to infinities, there is also no end to approaching absolute zero without never reaching it.

I think this is the most sound and simple explanation to your question.

  • $\begingroup$ how is the relation $\beta=1/(k_BT)$ regardless of $T$ sign? $\endgroup$ – Nikos M. May 30 '14 at 22:33
  • $\begingroup$ The point is that with this definition, heat always flows from small beta to high ones regardless of the sign. This encompasses also the weird fact that negative temperatures are hotter than infinite temoeratures. $\endgroup$ – gatsu May 31 '14 at 7:41

Quantum mechanics and the uncertainty principle (which just answered a while ago) states that consituting particles of a system have to undergo movement (change of state), else they could be localized with arbitrary precision.

Absolute zero temperature by (classical) definition is the temperature where a system is a "perfect crystal" (the energy state of a certain symmetry and nothing moving or changing further).

So combing the above 2, the result follows that absolute zero is asymptotical.

Quantum mechanically, even the ground state of the "void" is not at zero temperature, but can indeed exhibit fluctuations.


Check this question regarding a further connection between quantum uncertainty principle and thermodynamics.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy