Methods of methods in use for the production of low temperatures are as follows: 1. Freezing Mixtures—The most common freezing mixture, ice and salt, produces when mixed in suitable proportions (1 part salt to 3 parts ice) a temperature of —23° C. In 1834 Thilorier found that by mixing solid carbon dioxide with ordinary ether he could obtain a temperature as low as —110° C. This °Thi lorier's Mixture* was for a long time the only means available for securing very low tem peratures.
2. Evaporation.—The first use of this method of cooling was in 1824 by Bussy, who found that he could get a temperature as low as —65° C. by rapidly evaporating liquid sul phur dioxide. By this means he liquefied chlo rine, ammonia and even solidified cyanogen. This method of refrigeration is still in use. Subsequent to Bussy's experiments it was soon learned that both carbon dioxide (CO,) and ammonia (NH,) are well adapted to liquefac tion with subsequent evaporation to produce low temperatures. These two gases are in com mon use in our present-day refrigerating sys terns, carbon dioxide being preferred where leaks due to accident would make the ammonia system undesirable or even hazardous, as, for instance, on board ship.
3. Expansion (without external work).— The first practical use of the fact that a gas undergoes a drop in temperature when allowed to expand through a small orifice, without do ing external work, was made by Thilorier in 1834, when he liquefied and solidified carbon dioxide by this method.
In 1799 Van Marum observed that at high pressures ammonia gas does not obey Boyle's Law (i.e., that the product of the pressure and volume of a gas is a constant quantity). An extended series of tests by Amagat, extending up until 1888, has shown that the same is true of practically all gases, i.e., under high pres sures, the volumes of all gases (except hydro gen, helium and neon at comparatively high temperatures) are less than those calculated by the gas laws, due to the tendency of the gas molecules to cohere as the approach is made to the liquid state. A °perfect° gas expanding into a vacuum should undergo no fall in tem perature, since it would do no work, but ordi nary gases do become cooled.
The work they do in expanding thus con sists in overcoming the cohesion between their molecules, so that a tearing apart of the sub stance occurs and this consumes heat; this action takes place so quickly that there is no time to absorb heat from surrounding ma terials, and as a result the gas itself is cooled. Since the cohesion of the molecules becomes more marked as the temperature approaches the liquefaction point of the gas, the cooling effect of expansion becomes also greater as the tern perapre falls, The heat lost by the expansion of non-ideal gases was investigated in 1854 by Joule and Thomson. The cooling of gases in this way is
known as the Joule-Thomson effect, and it holds good for all gases except hydrogen (helium and neon) which above —80° C. is warmed by expansion.
The Linde apparatus (1898) for the lique faction of air depends on the Joule-Thomson effect. In this apparatus the gas under 200 atmospheres pressure is precooled by an ordi nary refrigeration machine and introduced into the actual liquefying column, where it is allowed to expand through an orifice without performing external work; the gas, cooled thus, is used to cool more incoming gas which in turn expands through the orifice and becomes further cooled; thus the temperature falls lower and lower until the liquefaction point of the gas in question is reached. Suitable ar rangements are made for withdrawing the liquid as desired, or, in the case of air lique faction, for submitting the liquid mixture to fractional distillation, by which method oxygen, nitrogen, argon, etc., may be obtained. Obvi ously, hydrogen cannot be liquefied in such an apparatus unless previously• cooled lower than —80° C., as was shown by Dewar, Ramsay, Olszewski and others.
4. Heat Exchanger.— The so-called °Regen erative Method° of cooling refers merely to the use of a gas cooled by expansion to further cool more incoming gas, thus obtaining further decreases in temperature. This method was first used in 1857 by Siemons and was applied to liquefaction of gases at the suggestion of Houston in 1874. This principle is embodied in all the liquefaction processes, the arrange ment being known as a aheat-exchanger," a series of concentric tubes in which the less cool gas passes along the inner tube counter cur rent to the cooler gas as it flows in the sur rounding outer tube. So efficient are well constructed heat exchangers that in manufac turing liquid air there is a difference of only 3 or 4° C. between the temperatures of the entering air and the escaping gases; i.e., dur ing the passage of a few seconds through the exchanger, the escaping air is heated from —190° C. to nearly room temperature. Obvi ously the manufacture of liquid air owes much of its success to the aheat-exchanger." 5. Expansion (with external work).— In this case an expansion engine is inserted in the circuit of a liquefaction system so that the gas expanding through the small valve (ori fice) may do external work and thereby lower the temperature more than can be accomplished by mere expansion without external work. This principle was applied commercially by Claude and his work has made possible an industry that has expanded until at the present time liquefaction machines based on the Claude system are located in all parts of the world for the production of oxygen and nitrogen from atmospheric air.