Whenever I wash my thermos, I put hot water and then some soap in; then I seal the one end with my hand or use the lid. After shaking it up, if I slowly remove the lid or my hand, it expels a little air. Why is that? Does it have something to do with increased surface area of soapy water? Or is it the fact that the air is heated by the water, even though the water must surely cool slightly?

  • 14
    $\begingroup$ I am so glad somebody asked this question. I have observed the same thing. $\endgroup$
    – M. Enns
    Feb 25, 2021 at 15:53
  • 17
    $\begingroup$ FWIW: In my on experience, you don't need the soap. Shaking a container in which, just moments earlier, some clean, almost-but-not-quite boiling water was sealed will increase the pressure in the container. $\endgroup$ Feb 25, 2021 at 17:11
  • $\begingroup$ It's true the effect works fine without soap. I have also managed to measure that if you shake the container with cold water the pressure drops which supports Gert's answer. $\endgroup$
    – M. Enns
    Mar 18, 2021 at 15:23

2 Answers 2


When you pour the hot water in, the air inside the thermos is still quite cold (ambient temperature, approx.)

But then when you shake it up the cold air is heated by the hot liquid. Gases expand considerably when heated, approximately acc. the Ideal Gas Law:


This causes a modest (and harmless) pressure increase in the flask, which is what you experience.

  • 21
    $\begingroup$ You can also observe the opposite effect when trying to open a fridge again right after closing it. $\endgroup$
    – noah
    Feb 25, 2021 at 16:42
  • 4
    $\begingroup$ @noah Why "right after closing it"? Shouldn't the air pressure difference become larger if you give the air in the fridge more time to cool down? Is it because fridges are too leaky to allow pressure differentials to persist? $\endgroup$
    – Will
    Feb 26, 2021 at 0:16
  • 3
    $\begingroup$ @Will yeah that's right. The cooling down will create a low pressure inside, but this is then slowly equalized by air leaking back in (which is cooled down again, maintaining a little bit of pressure difference). $\endgroup$
    – noah
    Feb 26, 2021 at 7:34
  • 5
    $\begingroup$ Funnily enough, a (very slightly) leaky freezer seems to be a feature not a bug. You'd never get the door open otherwise!! $\endgroup$
    – josh
    Feb 26, 2021 at 9:49
  • 3
    $\begingroup$ @SelfEvident The increase in temperature you can impart that way is really small. See Joule's experiment. $\endgroup$
    – Gert
    Feb 26, 2021 at 19:23

There is another effect here which is significant, as follows.

Warm water wants to evaporate, but in a flask-shaped container, the evaporation can take place only at the free surface of the water in the flask. Furthermore, as soon as the boundary layer of air right next to the warm water becomes saturated with vapor, the diffusion of water vapor into the air slows down greatly.

If you close the container and shake it vigorously with soap added, tiny air bubbles get mixed into the warm water, producing a huge surface area available for evaporation to occur across. The bubbles expand as they get loaded with vapor and the pressure inside the container jumps up suddenly.

  • 3
    $\begingroup$ +1 This effect can also be observed with bottles of cold water and soap. $\endgroup$
    – Sanchises
    Feb 26, 2021 at 13:10
  • 2
    $\begingroup$ Does this process/effect have a name? do you have a published reference on it? Not doubting you, but i'd like to explore this further and would need a google-entrypoint $\endgroup$
    – bukwyrm
    Feb 28, 2021 at 14:09
  • 1
    $\begingroup$ try "evaporation kinetics" $\endgroup$ Feb 28, 2021 at 22:02

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.