Dry air is made up of approximately 78% Nitrogen, 21% Oxygen, and 1% Argon, by volume. There are trace amounts of other stuff, like CO2, Neon, and Helium. Liquid Nitrogen is fun stuff to mess around with; even better, liquid Oxygen is both fun and dangerous! Who could ask for more?
Air also contains 1% to 4% water vapor as well.
Air was first liquefied by Carl von Linde. He used a very clever trick to use the use air cooled by expansion to pre-cool itself - a feedback effect that produced extremely low temperatures.
In the above image, air is compressed (the compressor is the circle), increasing its temperature (red). The hot, high-pressure air is first cooled to near ambient (mauve) by ground water in a heat exchanger (left-most square). The ambient-temperature high-pressure air then further cooled by a second heat exchanger (middle square) to below ambient temperature (blue). The cool, high-pressure air is then allowed to expand through an expansion nozzle, which causes it to become very cold (pale blue), into an insulated collection vessel (right-most square).
So far, it's just a refrigerator. Von Linde's trick, however, was to pipe the cold ambient-pressure air from the collection vessel (or at least that portion of it that didn't liquefy) back to the second heat exchanger (middle square) and use it to cool the ambient-temperature high-pressure air coming from the first heat exchanger. The more the high-pressure air is cooled before expansion, the colder the ambient-pressure air will be after expansion, further cooling the high-pressure air, making the ambient-pressure air even colder... positive feedback. The longer it runs, the colder it gets. Very clever.
Since the second heat exchanger isn't 100% efficient, the returning air coming out of it will still be below ambient. Rather than waste the effort put into extracting the heat from this air, it is fed back to the intake of the compressor and re-compressed. A little ambient air is added as the volume air in the system decreases from cooling and liquefaction.
Someday I am going to try to do this. I figure that an automotive air conditioning condensers will work well for the heat exchanger, the first dropped into bucket of water, the second in a Styrofoam cooler (I have one from Omaha Steaks that should work quite well). I've got a nice stainless steel thermos (to which I've lost the lid, so it's fair game!) that will work for a collection vessel.
There would need to be a trap for water condensation between the first and second heat exchangers. The second heat exchanger will probably fill up with dry ice - not sure what to do about that.
It occurrs to me that if the device can only reach a temperature between that of liquid Oxygen (-279°F) and liquid Nitrogen (-320°F), the Oxygen will liquify, leaving the Nitrogen as a gas. The device will soon fill up with pure Nitrogen and stop functioning.
I may be re-inventing the wheel here (it's what I do best!) but it seems to be that by carefully monitoring and controlling the temperature in the collection vessel by dumping some of the recycled air before it returns to the second heat exchanger, pure liquid Oxygen could be collected. Dumping the cooled air is necessary to keep the system from filling up with Nitrogen, as mentioned above.
By controlling the temperature to between that of liquid Argon (-309°F) and liquid liquid Nitrogen (-320°F) would be a bit tougher, but one could then collect a mixture of Oxygen and Argon. Warming that mixture to above -309°F would separate the Argon, which could be collected. Liquid Nitrogen could then be used to re-liquify it. Think "Fractional Distillation".
If you have to ask, you're not entitled to know.
© 2005 W. E. Johns