ALUMINIUM - ISOLATION AND EXTRACTION When in the middle of the i 8th century it was recognized that alumina was the oxide of a metal, all means available at the time were employed, though unsuccessfully, to smelt it. Indeed, La voisier, in 1782, expressed the belief that alumina could not be reduced by carbon, a belief maintained until, almost a century later, the brothers E. H. and A. H. Cowles, heating alumina mixcd with other metal oxides and carbon by electricity, produced aluminium alloys commercially. They failed, however, to win the pure metal and not until 19o7 was it shown that alumina alone could be reduced by carbon though only at a temperature (2,200° C.) at which the metal produced was volatilized and lost or recombined with the oxides of carbon formed. At the same time the function of the alloying metal used by them was found to be the prevention of this recombination and the absorption of the aluminium vapours set free.
Returning to 1809, we find Sir Humphry Davy passing an elec tric current derived from a voltaic pile through fused aluminium into iron molten in an atmosphere of hydrogen. Although there resulted an alloy of aluminium and iron in place of the metal he had hoped to isolate, Davy had thus accomplished the first and, perhaps, the greatest step toward the production of aluminium, for he had definitely proved that its oxide was reducible.
Abandoning electricity, which had failed Davy, H. C. Oersted and after him F. Wohler, attempted the displacement of alumin ium from its chloride by the alkali metal, potassium. It is un certain whether Oersted succeeded: it is certain that in 1827 Wohler produced metallic aluminium as a grey powder and, subsequently, in metallic pellets. Twenty-five years later, F. Rose, substituting, for volatile aluminium chloride, the newly imported and more stable double fluoride, cryolite, and H. St. Clair Deville, replacing potassium by the cheaper sodium, converted Wohler's laboratory method into successful technical processes, which were practised in several countries until both were alike swept from the field when the evolution of the dynamo made it possible to produce aluminium electrolytically.
A method, based on the power of electricity to split metal salts into their components, had been discovered many years before it was possible to apply it commercially. Earlier attempts to elec trolyse aqueous solutions had failed because, as we now know, the aluminium produced at once recombined with water. In 18S4, however, Robert Bunsen passed a current through a molten mixture of aluminium chloride and common salt, splitting the for mer into chlorine, which escaped, and aluminium which he col lected as metallic beads. Bunsen's method was developed 3o years later into the process by which aluminium is produced to day. Once more cryolite replaced the volatile chlorides and this led to a further step of the utmost importance. It was found that molten cryolite readily dissolves alumina and that a current passed through such a solution decomposes, not the cryolite, but the cheaper and more easily resolved alumina, thus economizing both power and material, and, as will be shown, making possible a continuous process. It was Paul Heroult in France and Charles Hall in America who, in the '8os made this great advance, the Frenchman as the result of an accident, the significance of which he had the wit to recognize and the energy to turn to account; the American as the result of a definite quest doggedly pursued.
Since cryolite is excessively corrosive the operation is carried out in open-topped iron casings, thickly lined with carbon. The cur rent is led into the furnace by a number of carbon blocks sus pended in the bath and leaves through the furnace lining. Alumina is decomposed into aluminium and oxygen and the latter consumes the carbon anode blocks, forming gaseous oxides of carbon (see fig. I). The operation is carried on at about 95o° C. or far above the melting point of aluminium (659° C.) which is, therefore, liquid when produced. Being heavier than the electrolyte it sinks to the bottom of the bath, the current passing through it to the lining, thence to the furnace casing and away. As the alumina is consumed fresh quantities are added and as the carbon anodes burn away they are replaced, whilst the aluminium is tapped from the furnace periodically, so that the process is practically con tinuous. The baths are about 'ft. deep, aft. long and 21 to Sift. wide. Heavy currents of from 8,000 to 20,000 amps. are used, the voltage for each bath, of which 4o-5o are placed in series, being about 5.5-6.5. Theoretically, the voltage should be only 2.79 and ro,000 amps. should produce about 177 lb. of aluminium per day. In practice this becomes about i5o lb. and it takes to produce ton of aluminium per year. Theory de mands 1.88 tons of alumina per ton of aluminium and this figure is but very slightly exceeded, whilst about ton of carbon is consumed. Theoretically no cryolite is used but in practice there are losses amounting to about o• 1 ton per ton of aluminium.