Why does oxygen not condense onto open dewers of liquid nitrogen?

I had a discussion with my supervisor today after I was a touch careless pumping down a flask with a liquid nitrogen cold trap attached. I checked, and there are reports of explosions after running air through a liquid nitrogen trap.

Given this, why does liquid oxygen not condense when you leave an open dewer of liquid nitrogen out? I mean, the air touching the surface of the nitrogen on top of the trap is the same as the air pumping through it. Is there too much nitrogen evaporating for the air to get close, or is the surface of the nitrogen too warm, or some other reasons?

• There's an additional factor in a liquid nitrogen trap that makes it dangerous, and that is that you often condense organic solvents in it, which can be ignited by the liquid oxygen. Commented Jul 13, 2012 at 9:59
• There is a worrying tendency in answers here to explain why something doesn't happen when, in fact, it does happen. Commented Jul 3, 2015 at 22:02
• @matt_black We leave our glovebox cold trap up for hours at a time, and I've never seen any blue when pouring it out. Commented Jul 4, 2015 at 23:49

The boiling point of nitrogen is 77 K, and the boiling point of oxygen is 90 K, so oxygen is liquid when liquid nitrogen is around. Wikipedia repeats the warning you mentioned:

Liquid nitrogen has a lower boiling point at −196 °C (77 K) than oxygen's −183 °C (90 K), and vessels containing liquid nitrogen can condense oxygen from air: when most of the nitrogen has evaporated from such a vessel there is a risk that liquid oxygen remaining can react violently with organic material.

So it's not because the surface of the nitrogen is too warm.

My guess is that the pressure due to the evaporating nitrogen is pushing air away from the opening. LN2 has an expansion factor of 694, which you can compute by dividing the density of the liquid (0.808 g/mL) by the density of the gas at 20 °C (1.17 g/L). Of that expansion factor, a factor of 303 K/77 K = 3.94 comes from warming the gas from 77 K to 20 °C. The other factor of 176 comes from turning the liquid into a gas. That expansion is pushing air out of the headspace of the dewar.

• Even if a little liquid oxygen does condense in the dewer, the partial pressure of oxygen will still be lower than normal air unless significant quantities of oxygen condense, which seems unlikely given the expansion. There is also typically nothing combustible inside a dewer, unlike a trap which you just condensed solvent in!
– Nick
Commented Jul 13, 2012 at 15:27
• The fact that liquid oxygen can exist at 77K doesn't mean that oxygen at 77K and a partial pressure of 0.21bar will condense. In fact, your liquid oxygen at 77K will (very slowly) evaporate if I'm reading that vapour pressure graph correctly. Commented Jul 18, 2012 at 2:45
• @BerndJendrissek, my understanding of that vapor pressure graph is that it is the total pressure that is relevant, not the partial pressure of oxygen. As a comparison, water boils when its vapor pressure reaches 1 bar. Commented Jul 18, 2012 at 20:03
• @ColinMcFaul Pure water boils when its vapor pressure reaches 1 bar. Dilute it with, for instance, ethanol, and the boiling point is depressed. Commented Apr 12, 2013 at 22:28
• The boiling point occurs when the vapor pressure equals the ambient total pressure. That's why the BP of water decreases with elevation- lower atmospheric pressure. The boiling temperature doesn't change from day to day depending on how humid it is. Commented Apr 18, 2013 at 7:00

If you look at the vapour pressure graph for oxygen you’ll see that at its partial pressure in air, it can only liquefy below about 77 K, which by coincidence is the temperature of boiling liquid nitrogen. The air immediately above the liquid nitrogen / air interface cannot be any colder than that 77 K (slightly less, because nitrogen is only 4/5ths of the air), so the 1/5th that is oxygen cannot actually condense – it doesn’t get cold enough. Adding to that is the fact that the evaporating nitrogen further dilutes the oxygen that is present, further lowering the temperature that would be needed to condense oxygen at that pressure.

• At atmospheric pressure, oxygen liquifies at 90 K, not 77 K. Commented Jul 17, 2012 at 17:29
• @ColinMcFaul, where do you find oxygen at 1 bar other than in, say, the Apollo 1 capsule? If you think of oxygen as "water", then just above the liquid nitrogen surface, where it is just a tad under 77K, the "relative humidity" is at most (thumbsuck) 97% - still under 100%. That means it can't "rain" liquid oxygen. Since evaporating nitrogen dilutes what oxygen there is, the air might be even "drier" than that 97%. Even less chance of oxygen rain. Commented Jul 18, 2012 at 2:41
• Why does oxygen condense in vapour traps then? Commented Apr 25, 2013 at 1:04
• The link is broken... Commented Jan 10, 2017 at 13:43

Actually, it does.

Normally, though, you won't notice it as it will be heavily diluted by the nitrogen and won't ever reach a concentration large enough to worry about. It's different if you pull air through a trap as the oxygen content of the flowing gas will be condensed preferentially leaving a liquid that is concentrated in oxygen. This is really dangerous if there is any organic material in the trap as well which is why those of who have worked with such systems are so wary of pale blue liquids in our traps.

But, if you like dangerous experiments, you can do some that illustrate the presence of oxygen. If you cut of the top of an aluminium can and half fill it with liquid nitrogen, then wait until most has evaporated, the remaining liquid will be mostly oxygen. The safe way to show this is to use a magnet on a stick to pull some out (oxygen is fairly strongly paramagnetic). Or, much less safely, you could drop a lit match in (this is very dangerous as the can will probably burn violently). There is even a youtube demonstration of this (I think it is number 5 of 7 "super cool" demonstrations with liquid nitrogen). I think they used to have a version with burning but may have taken it off for safety reasons.

Oxygen can and does condense into liquid nitrogen dewars. Good lab practice dumps vacuum line dewars Friday afternoon, refills and restarts Monday morning. Remember – the universe hates you.

Dumping old liquid nitrogen in the parking lot can be fun. LOX plus asphalt paving (or wooden floors) is a contact explosive for days. Charcoal briquettes plus LOX was a blasting combo in coal mines. NEVER pour LOX on burning charcoal briquettes. The fireball burns so hot you can get UV burns, assuming you have any skin remaining. Steel wool plus LOX is a thermal lance, minimum 2730 °C, sometimes claimed to be 4500 °C.

Take care that anything you wear when working with cryogens can be removed very fast – especially porous insulating gloves.

Vapor pressure is property of liquid and partial pressure is property of a gas in a mixture.

The normal boiling point for the pure component in liquid phase is the temperature at which vapor pressure exerted by the liquid becomes equal to 1 atm.

In a gas mixture, as long as partial pressure of a component remains less than its liquid phase vapor pressure at the temperature of the mixture, the component will not condense. Condensation starts only from the point when $PP=VP$ and starting from the case where $PP>VP$, condensation continues only till $PP=VP$. When $PP<VP$, condensation stops.

Air at 1 bar total pressure will have 0.21 bar partial pressure of $\ce{O2}$ and say it is at -183C (the normal boiling point of $\ce{O2}$). Vapor pressure of liquid $\ce{O2}$ at this temperature is 1.01 bar, but partial pressure is 0.21 bar which is less than 1.01 bar so $\ce{O2}$ should not condense. But if Air is at 5 bar total pressure, this will have $0.21\times 5=1.05$ bar partial pressure of $\ce{O2}$ and this air is at -183C then $\ce{O2}$ will condense till its partial pressure has gone below 1.01 bar.

Vapor pressure is independent of composition in liquid phase but dependent on temperature. As temperature increases, the vapor pressure of a liquid component increases. Vapor pressure does not depend on gas phase mixture properties. At a given temperature, the component having the higher vapour pressure in liquid phase is the more volatile components and will have lower boiling point.

• Welcome to Chem SE. I am guessing your answer is in response to one of the other answers. Typically, that activity should be in the form of a comment instead of another answer. You no doubt posted it as an answer because you don't have enough reputation to comment. It only takes 50 rep to be able to comment (on others' posts), which you can get to pretty quickly by asking, answering and editing. Each vote up is 10 rep, so you'd only need to get voted up five times. Good luck! Commented Apr 18, 2013 at 7:19