I'd like to do some experiments with liquid oxygen but as a first step I'll have to make some. Unfortunately I don't have any experience designing cryogenic experiments. I'm thinking I'll cool gaseous oxygen using liquid nitrogen as the coolant and then I want to collect the liquid oxygen condensate in a small dewar of some kind.

It needs to be high purity so I'll be using 99.9% purity "scientific grade" oxygen from a cylinder.

Does it make sense to use conventional condenser glassware like a dimroth condenser for instance? Alternatively what's the best equipment to use and what's the best way to rig a cryogenic setup like this?

Also what safety precautions should I take?

  • $\begingroup$ You can probably just condense air with liquid nitrogen in a dewar, but it is dangerous. $\endgroup$ Sep 28, 2015 at 0:35
  • $\begingroup$ Ah, I forgot to mention that this has to be pure oxygen, not liquid air. $\endgroup$ Sep 28, 2015 at 1:52
  • 1
    $\begingroup$ Fortunately for safety, oxygen boils at a lower temperature than nitrogen. Fractional distillation isn't 'hard', but complex at cryo temperatures. $\endgroup$
    – Jon Custer
    Sep 28, 2015 at 2:16
  • 1
    $\begingroup$ @JonCuster That is incorrect. The boiling point of nitrogen is about -195 degree Celsius and that of oxygen is about 183 degree Celsius. $\endgroup$ Sep 28, 2015 at 2:19
  • 1
    $\begingroup$ You can use something as simple as a test tube immersed in liquid nitrogen (in a Dewar) that is charged with oxygen from tubing connected to your oxygen cylinder. But this will be open in the air. Otherwise any closed system I imagine it has to be specialised equipment since any oxygen evaporation can create high internal pressure in a closed system. (So a Schlenk tube would not be a good idea) $\endgroup$
    – K_P
    Sep 28, 2015 at 7:43

2 Answers 2


Simply put an open flask in the LN2 and it will fill with liquid oxygen from the air.

Or better: DON'T DO THAT!! What the heck do you want with it?!?

LO2 is highly dangerous. If it comes in contact with anything combustible, it will blow up vigorously. Together with lab solvents, it will likely explode. Make sure the flask or dewar is dead clean. Chemistry is science, and science is about prediction. If you cannot predict, from knowledge, not hearsay, what your experiment will likely do, don't do it.

  • 1
    $\begingroup$ I really think this is the sort of question where if you have to ask then you don't understand the situation enough to do the experiment safely. Safety can't be an afterthought but must be planned into the experiment. $\endgroup$
    – MaxW
    Jan 10, 2016 at 3:46

It has already been suggested to use liquid nitrogen to cool oxygen and that is probably your best bet. You could research something called a cryocooler that can be used to condense air or pure oxygen but it is very costly and would not be worth it unless you plan on using a continuous supply liquid oxygen and/or nitrogen. If you're just going to do few experiments, don't worry about the cryocooler idea.

But I will say this. You can condense air from LN2 and end up with very pure LOX. Immersing an empty test tube in a container of LN2 for a while, you will notice it slowly filling with a very very pale blue liquid. But if you remove the tube from the LN2 and let it boil for a while, you'll notice it getting bluer and bluer. Once it gets a little darker than sky blue, you can be sure you have nearly PURE liquid oxygen remaining.


Your Answer

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

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