When and how did air separation start?

 In May 1895, Carl von Linde performed an experiment in his laboratory in Munich that led to his invention of the first continuous process for the liquefaction of air based on the Joule-Thomson refrigeration effect and the principle of countercurrent heat exchange. This marked the breakthrough for cryogenic air separation.

For his experiment, air was compressed from 20 bar [p₁] [t₄] to 60 bar [p₂] [t₅] in the compressor and cooled in the water cooler to ambient temperature [t₁]. The precooledair was fed into the countercurrent heat exchanger, further cooled down [t₂] and expanded in the expansion valve (Joule-Thomson valve) [p₁] to liquefaction temperature [t₃]. The gaseous content of the air was then warmed up again [t₄] in the heat exchanger and fed into the suction side of the compressor [p₁]. The hourly yield from this experiment was approx. three litres of liquid air. Linde based his experiment on findings discovered by J. P. Joule and W. Thomson (1852). They found that compressed air expanded in a valve cooled down by approx. 0.25°C with each bar of pressure drop. This proved that real gases do not follow the Boyle-Mariotte principle, according to which no temperature decrease is to be expected from expansion. An explanation for this effect was given by J. K. van der Waals (1873), who discovered that the molecules in compressed gases are no longer freely movable and the interaction among them leads to a temperature decrease after decompression.