Thermometry

It is often convenient to repre sent the behaviour of mixtures by means of a temperature-con centration diagram at constant pressure, as has been done by E. C. C. Baly for air (fig. 23). At any temperature the vapour contains more nitrogen than oxygen.

Oxygen from Liquid Air.

The most remarkable result of investigations at very low temperatures has been the development of the process for the separation of oxygen from air. The boiling point of oxygen is —183.0° while that of nitrogen is —195.8° C, so that unless the liquids formed constant boiling mixtures, which is not the case, separation of the gases by liquefaction and sub sequent rectification of the liquid presents no difficulties. The separation was first effected successfully by Linde on a technical scale in 1901. He adopted a method long in use by chemists for the separation of volatile liquids by distillation. Liquid air was allowed to trickle down a vertical cylinder, filled with glass balls, into a chamber at the bottom in which it was partially evaporated by cooled compressed air, passing through a coil immersed in it. The vapour passed upwards, meeting the descending stream of liquid, which washed out the oxygen, so that the gas leaving the top of the cylinder was mainly nitrogen, with only 7% of oxy gen, while liquid containing 96 to 98% of oxygen collected at the bottom. This liquid was drawn off and evaporated, and in this way 70% of the oxygen of the air could be separated with an impurity content of only 2% to 4%. Small oxygen plants of this simple type are in general use, but they are inefficient, as about one third of the oxygen is lost.

Much more efficient separation is effected in plants operated on a principle first introduced by Claude in which the gas is frac tionated in two stages, the first operated at about four atmos pheres, and the second at atmos pheric pressure. The working of the system is illustrated by the diagram (fig. 24). The diagram does not include the heat ex changer, liquefier, or expansion engine. Air under a pressure of 33 atmospheres passes first through one of a pair of heat ex changers, around which, alter nately, the cold separated nitro gen and oxygen from the plant pass, and in it the moisture from the air is condensed. Through the heat exchanger which is not being operated, the compressed air first passes, and thaws out any ice which had separated while the cold gas was passing. In the heat exchanger the temperature of the compressed air is reduced to 100° C. From it 6o to 70% of the compressed air passes to the expansion engine, in which it expands to four atmospheres doing external work, and being cooled to near the point of lique faction. The remaining 3o% to 4o% of the air passes at full pressure into a liquefier, over the coils of which the cold separated oxygen and nitrogen pass direct from the separator, before enter ing the heat exchanger which is in use. The liquid air from the liquefier and the cold air from the expansion engine now enter the outer chamber A at the bottom of the separator through the pipes Q and P respectively. Here, since we have at the first moment of condensation a relatively small quantity of liquid in contact with a large quantity of vapour a condition of equilibrium exists corresponding to that at the moment when air begins to liquefy, when we have a vapour phase con taining rather less than 21% of oxygen, which is in equilibrium with a liquid containing rather less than 47% of oxygen. The air

which is not liquefied passes upwards through the vertical tubes B which traverse the chambers C and D which, as we shall see later, contain nearly pure liquid oxygen at about —183° C, and at nearly atmospheric pressure. The air, under nearly four at mospheres pressure, partially liquefies in contact with the walls of these tubes, and the liquid running back down the tubes, washes the oxygen out of the ascending stream of air, the liquid collecting in A having an average concentration of 4o% oxygen. The gas, now poor in oxygen passes, from the top of the outer nest of tubes, down an inner set of tubes where most of it is condensed, its oxygen content being about 4%, and runs into the inner chamber E. The greater the quantity of the oxygen-nitrogen mixture which is condensed in A, the lower will be its oxygen content, and the richer in nitrogen will be the liquid which passes into E. Claude describes this part of the process in terms which may be rendered as antecedent liquefaction and separation by backward flow. The essential feature of the process is the con densation at a pressure above atmospheric pressure, and utilising the heat of condensation to evaporate the liquid in the cycle oper ated at atmospheric pressure. The principle is similar to that of the double effect still. As the pressure in that part of the appara tus, which includes the chamber A, the ascending tubes B, and the tubes descending to E, is four atmospheres, the liquids con densed in A and E can be raised to the top of the apparatus and delivered into the fractionating column, which is similar to the column in a spirit still, the temperature at the top being not far above C, the boiling point of nitrogen, and that at the bottom approaching —583° C, the boiling point of oxygen. The nitrogen rich liquid from E is delivered to the top of the column at H, and the oxygen content of the liquid at this point is in creased, and the oxygen content of the vapour consequently diminished by the upward passage of the gas rising through the column. The liquid from A is delivered into the column at a point K at which the composition of the liquid on a tray corre sponds to that of the liquid forced up from below. The liquid descending through the column is scrubbed by the gases produced by the evaporation of the liquid in D, so that the liquid which is partly evaporated, and partly passes as liquid into C, where it is completely evaporated, and the gas which is 99% oxygen, the impurity being mainly argon with traces of krypton and zenon, pass through G to cool the liquefier and heat exchanger, and thence to the gas holder. The nitrogen from the top of the ap paratus is similarly used to cool the incoming air. A small set of tubes in the middle of the lower part of the apparatus allows of a small quantity of nitrogen rich gas nearly free from oxygen, but containing the whole of the helium and neon in the air, being separated and drawn off through the pipe R. Oxygen plants of this type are producing 4,00o cu.ft. of oxygen per hour at an expenditure of 46 B.H.P. per i,000 cu.ft. of oxygen.