Cookies   I display ads to cover the expenses. See the privacy policy for more information. You can keep or reject the ads.

Video thumbnail
If you’ve ever used a can of compressed air (also called a gas duster), to, say, clean
crumbs out of your computer keyboard , you’re probably aware that after a little while,
the air coming out of the can and even the can itself get really really cold.
Like, cold enough they put frostbite warnings on the can!
And for good reason!
It’s tempting to think that compressed air cans get cold because when the gas comes out
of the can it expands and thus cools off.
But that’s not exactly right - whether an expanding gas gets hotter or colder (and how
much hotter or colder it gets) depends on the exact manner in which the gas expands.
And if we apply the relevant equation for “normal” gas expansion , we predict that
the gas inside the compressed air can should drop from room temperature to around 100 degrees
celsius below zero , which is, um, WAY colder than what comes out of a compressed air can.
So the gas can’t be expanding in the normal way gases expand.
. And here’s why: that would be like cutting the top off the can and letting the gas expand
freely in all directions.
But the gas is actually being squeezed out through a tiny valve.
This difference is key; the gas passing through a valve isn’t simply expanding - it’s
also being pushed through by the rest of the gas behind it!
And that compression from behind gives the gas enough heat energy to essentially counteract
the cooling from expansion.
In terms of the gas law, this means the volume goes up by the same factor that the pressure
goes down, so pressure times volume is pretty much constant and the temperature stays about
But not exactly - most gases at room temperature do get slightly colder when passing through
a valve . A good demo of this is to let the air out of a bike tire; the valve gets colder
, but not crazy cold.
Similarly, the gas leaving a can of compressed air cools a little bit passing through the
nozzle . But this can’t be the only contributor to the cooling.
I mean, the can itself cools off by significantly more than can be explained by expansion through
a valve , and it’s not like it’s even being sprayed by the air coming out.
No, the real cooling power is hinted at by the warning labels on cans of compressed air
telling you to not to shake them or spray them upside down - if you DO shake one, you’ll
realize right away that it’s not just gas inside - there’s liquid in there, too!
Liquid like 1,1-difluoroethane, which is a gas at normal temperatures and pressures,
but a liquid once you pressurize it to around 6 times atmospheric pressure.
And it’s the essential component of these compressed air cans.
Inside the can, 1,1-difluoroethane exists as both a liquid and a gas, in equilibrium
- just enough of the liquid boils off to maintain six atmospheres of pressure in the top of
the can, a pressure high enough that rest stays liquid.
Because it’s at six times atmospheric pressure, when you open the valve the difluoroethane
rushes out in a steady stream.
But this then means that the inside of the can is no longer pressurized enough to keep
the liquid from boiling - so more of it boils off until the gas reaches six atmospheres
of pressure again and a new equilibrium is reached with slightly less liquid in the can.
This is how the can is able to keep blowing a stream of consistent strength even when
mostly empty.
But more importantly to our temperature conundrum, changes from liquid phase to gas phase require
a TON of energy, and that energy has to come from somewhere.
Just like how the evaporation of sweat removes energy from your skin, cooling you off, inside
a can of compressed air, vaporization - aka boiling - is what steals energy from the liquid
and cools it off.
Spraying out 10% of the contents will cool the entire remainder of the can by around
20 degrees celsius!
If it seems counterintuitive that a boiling substance cools itself off, look no further
than the humble pressure cooker . Water normally boils above 100 degrees celsius, but by sealing
in steam, the pressure rises, enabling the water in the pot to remain a liquid well beyond
water’s normal boiling point - just like the difluoroethane in a can of compressed
And releasing water vapor out of the nozzle of a pressure cooker lowers the pressure inside,
allowing a bit more water to boil off as steam and lowering the temperature of the remaining
water - just like the difluoroethane in a compressed air can.
And if you keep letting off steam, eventually the water will cool all the way back down
to its regular boiling point of 100 degrees , just like how if you keep spraying a can
of compressed air, the difluoroethane inside will cool all the way back down to its regular
boiling point of negative 25 degrees .
A can of compressed air is quite literally a 1,1-difluoroethane pressure cooker.
And just like you shouldn’t shake a pressure cooker or turn it upside down (unless you
want to spray superheated water everywhere), cans of compressed air don’t work very well
sideways or upside down: instead of spraying out gas, you’ll spray out the liquid that
was only being kept liquified by the high pressure inside the can , so it immediately
vaporizes and drastically cools down whatever it’s contacting . INSTANT ICE! (though difluoroethane
can dissolve in water and is poisonous, so definitely don’t use this ice for anything
In conclusion, the cause for the coldness of cans of compressed air can be clarified
by comprehending the consequent clue: they aren’t actually cans of compressed air.
They’re cans of pressure-liquified 1,1-difluoroethane, and lowering the pressure inside by spraying
them allows more liquid to boil off, cooling what remains.
I love learning about the physics of regular stuff; I mean, black holes and quantum mechanics
are cool, too, but they’re not quite as tangible or relatable as the things we interact
with on a regular basis.
And if you, too, want to dive deeper into the physics of everyday objects, look no further
than Brilliant, this video’s sponsor.
Brilliant has a whole course on the physics of everyday objects, including fridges and
water towers and bikes - bikes are great! - and Brilliant also has fun, short daily
challenges and puzzles to learn about stuff like regression to the mean and fluids and
thermodynamics without the huge time commitment it would take to learn enough about Joule-Thompson
expansion through a valve to make a whole youtube video about it…
Brilliant continues to be an incredible supporter of MinutePhysics and they’re offering 20%
off of a premium subscription to the first 200 of you who go to,
which gives you full access to all of Brilliant’s courses, puzzles and daily challenges.
Again, that’s for 20% off a premium subscription, and to
let Brilliant know you came from here.