I've been searching the web for this phenomenon and surprisingly I found only one post about it. And the reply to this question about canned air begins with words:

I don't know about the shaking part (it is hard for me to see how that would make a difference)

This beginning has strongly shaken my faith, so for sure I picked up a can with deodorant and shaken it. And it does become cooler. Significantly! After standing still, the can doesn't seem to become warmer, so this reaction doesn't revert itself and maybe is irreversible.

As I imagine it, when being shaken the average pressure in the can as a whole can't possibly change. Also it's interesting how you add energy by shaking only to have more energy drained by cooling.

So please, can you explain this strange behavior? Could you also cover the energy flow? When does the energy return back out of the can?

  • $\begingroup$ OK! I admit it! You got me interested! Hmm, as the deodorant isn't very pressurized, adiabatic expansion upon releasing isn't playing an effective role here. But, if you have the bottle really close to your skin, you might be feeling the alkanes that instantly evaporate when sprayed. Shivering $\endgroup$
    – M.A.R.
    Commented Apr 8, 2015 at 19:29
  • 1
    $\begingroup$ I just tried this myself with a can in each hand simultaneously and only shook one, but I could not reproduce the effect to my satisfaction, even after some alternation. To me, if there is a difference, it is not large. Also, I can see some confounding factors, the main one being that we don't detect temperature, but rather the rate of heat exchange; the shaking hand will grip tighter, and also some heat is being generated by the hand while it is gripping and shaking, both of which favour removal of energy from the hand at a faster rate. $\endgroup$ Commented Apr 8, 2015 at 19:31
  • $\begingroup$ @NicolauSakerNeto You make good points. I tried to grip the can as hard as I can and I stil observed the effect. But I indeed need some measured evidence only I don't know how to get it. Also it doesn't seem to work with all cans. Definitelly does not work with bottles of water or glasses. According to your theory, I should observe the same effect even with empty glass. I'll also try to ask somebody unaware of the phenomenon without telling him what I expect them to observe. That's the only alternative to precise laboratory equipment which I don't have. $\endgroup$ Commented Apr 8, 2015 at 19:48
  • 1
    $\begingroup$ Here's another important factor; metal cans are quite thin and have small thermal capacity, which means most of the heat is transferred quickly and predominantly into the liquid, and from there on heat transfer is limited in part by how quickly warmer liquid molecules can diffuse away from the surface in contact with the hand. Shaking will allow cooler molecules to more quickly reach the contact region, once again speeding up transfer. If the container is made of a thermal insulator or has a high thermal capacity (thick walls), the liquid inside may make no difference in a short experiment. $\endgroup$ Commented Apr 8, 2015 at 20:04
  • $\begingroup$ @TomášZato Adiabatic expansion. $\endgroup$
    – user3735
    Commented Apr 10, 2015 at 0:39

2 Answers 2


TL;DR: Spray cans don't actually get colder when shaken. However, shaking a can does increase heat conduction from your hand to the can, making it feel colder.

Humans don't actually sense external temperature directly; our thermoreceptors are located under the skin, and thus effectively measure the rate at which body heat is lost through the skin. This is why, for example, touching a metal surface will feel noticeably colder than a wooden or plastic surface at the same ambient temperature.

When you pick up a spray can, it is typically at room temperature, i.e. significantly below body temperature. The can itself is a thin metal cylinder, and thus conducts heat well, but has very little heat capacity in itself. Thus, the rate of heat loss from your hands (and thus the sensation of coolness) mainly depends on how quickly whatever's on the inside of the can will absorb heat from the surface.

Now, a typical aerosol spray can contains a propellant (mixed with the actual payload fluid) which would be slightly above its boiling point (and thus in gaseous form) at room temperature and pressure, but is kept liquid by the elevated pressure inside the can. As the spray valve is opened and the pressure reduced, some of the propellant fluid will boil, bringing the pressure back up until the liquid and gas phases are again at equilibrium.

What this means is that, in a half-empty spray can, the top half of the can will contain gaseous propellant (which, being a gas, will not absorb heat very effectively), while the bottom half will contain liquid propellant (mixed with the payload fluid) at just under its boiling point. This liquid will absorb heat much better than the gas at the top, both because it's denser (and thus has higher heat conductance and capacity), and also because heating it will cause some of it to boil, converting part of the absorbed heat into its latent heat of vaporization. (The boiling effect actually causes a half-empty spray can to act, to some extent, as a diode heat pipe.)

Thus, even without shaking the can, it's possible to observe that the bottom end of a half-empty spray can will feel cooler to your hand than the top end.

Now, when you hold the can and shake it, you cause the propellant liquid to slosh around inside, coating all of the inner surface of the can instead of just the bottom. The sloshing also significantly increases the heat transfer from the can to the liquid, by causing the warmer liquid near the surface of the can to mix with the colder liquid in the interior. (This is essentially the same reason as why moving air feels cooler on your skin than still air.) Thus, a shaken can will generally feel even cooler to touch than either end of an unshaken can.


agitating the can causes the liquid / gas equilibrium to shift toward the gas phase. since the can is not a closed thermodynamic system, the required energy (latent heat of evaporation) is "borrowed" from the environment.

  • $\begingroup$ Why does agitation shift the equilibrium towards the gas phase? $\endgroup$
    – ron
    Commented Apr 14, 2015 at 22:39
  • $\begingroup$ Surface area is affected: dG = σ dA $\endgroup$ Commented Apr 15, 2015 at 1:07
  • $\begingroup$ this was merely a hypothesis, but seemed plausible enough $\endgroup$ Commented Apr 15, 2015 at 1:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.