Below is A Replica of the Aluminum Cap to the Washington Monument Located in Washington DC.
On February 23, 1886 in the woodshed behind his family's home at 64 East College Street in Oberlin, Charles Martin Hall produced globules of aluminum metal by the electrolysis of aluminum oxide dissolved in a cryolite-aluminum fluoride mixture. This achievement was the culmination of several years of intensive work on this problem. It was a goal that he had set as a high school student and that had been encouraged with ideas and materials by Professor Frank F. Jewett of Oberlin College. Hall was graduated from Oberlin in June 1885, eight months before the successful experiment.
At the time of Hall's discovery the only practical way to make aluminum metal was through the chemical reduction of anhydrous aluminum chloride by sodium metal at elevated temperatures.
This process was a costly one owing to the difficulty of preparing water-free aluminum chloride from aluminum oxide, the natural source of aluminum, and to the necessity of making sodium metal by chemical means.(1) In the early 1880's aluminum was a semiprecious metal, but Jewett had a sample of it to show his students.
Not only did Hall have to devise a method for winning aluminum metal, but he also had to fabricate most of his apparatus and prepare his chemicals. In his early experiments he tried to adapt the high temperature carbon reduction methods that were used in the metallurgy of iron and other metals of intermediate chemical activity. He also tried to reduce the aluminum in cryolite with sodium metal. In subsequent experiments Hall, working in Jewett's laboratory in Cabinet Hall, showed that the electrolysis of aluminum fluoride in water gave only aluminum hydroxide. They understood that electrolysis provided more powerful reduction conditions than did chemical methods. Today, it is well known that hydrogen in water is more easily reduced than is aluminum ion. The cathodic half reaction in the presence of aluminum ion is
Having shown that an aqueous system was useless for the preparation of aluminum by electrolysis, Hall turned his attention to the possibility of using water-free fused salts as solvents for aluminum oxide.(2) But first, he had to build a furnace capable of producing and sustaining higher temperatures than the coal-fired, bellows-driven furnace that he had used in earlier experiments. For this purpose he adapted a second-hand, gasoline-fired stove to heat the interior of a clay-lined iron tube.(3) Despite the high temperature in this furnace he was unable to melt the first substance he tried, the mineral fluorspar, which is calcium fluorite (m.p. 1360 degrees Celsius). He then synthesized and tried potassium fluoride (m.p. 846 degrees Celsius), sodium fluoride (m.p. 988 degrees Celsius), magnesium fluoride (m.p. 1266 degrees Celsius) and aluminum fluoride (sublimation point 1291 degrees Celsius). The potassium and sodium fluorides melted in the furnace but did not dissolve useful amounts of aluminum oxide. He was unable to fuse magnesium fluoride or aluminum fluoride.
Hall moved on to experiments with the double fluoride of sodium and aluminum, which was formulated 3NaF.AlF3 in his day and was known as the mineral cryolite. He knew that this material was available from natural sources. No doubt he also knew that mixtures of salts commonly had lower melting points than the higher melting of the two. Today, this substance is written with a formula of Na3AlF6. It is understood to be an ionic compound containing sodium ions and hexafluoroaluminate ions, AlF63-. Hall synthesized his cryolite. He also had to prepare aluminum oxide. He did so from alum, KAl(SO4)2.12H2O, which was a common household substance in his day. He dissolved alum in water, precipitated aluminum hydroxide by adding washing soda (Na2CO3), another common household substance, filtered off the hydroxide, and dried it. Hall's older sister, Julia Hall, who had studied chemistry in college, followed the experiments closely and probably helped prepare some of the aluminum oxide. He melted cryolite (m.p. 1000 degrees Celsius) in the furnace and quickly found that it was a good solvent for aluminum oxide. He did this signal experiment on February 9, 1886 and repeated it for his sister to see when she returned the next day from a visit to Cleveland.
To do electrolysis in the 1880's most people had to make batteries. Hall and Jewett made cells of zinc in dilute sulfuric acid and graphite in concentrated nitric acid. (Such cells, known as Bunsen batteries or Bunsen-Grove cells, were commonly used in electrolysis experiments before the advent of motor generators.) For Hall's experiments this was a large undertaking. Due in part to inefficiencies in the process, about 5 moles of zinc metal (~300 g) would be consumed in making one mole (~30 g) of aluminum metal. Though a good source of high current, the Bunsen battery emits noxious fumes of nitrogen oxides.(4)
Hall's first attempts to prepare aluminum metal by electrolysis were carried out on February 16, 1886. He used graphite-rod electrodes dipping into a solution of aluminum oxide in molten cryolite held in a clay crucible. Because he observed gas formation around the positive electrode (anode), Hall was confident that electrolysis was occurring. However, when Charles cooled the melt and broke it open in Julia's presence, they found only a grayish deposit on what had been the negative graphite electrode (cathode). He repeated this experiment several times over the next few days. Finally, he recognized that the new material, which did not have the shiny metallic properties of aluminum, was probably silicon, which is a metalloid. Suspecting that this silicon had its origin in the silicates in the clay, he decided to fabricate a graphite crucible to use as a liner for the clay crucible. This graphite crucible was only 2 inches wide and 4 inches deep.(5) The first electrolysis experiment with this system, in which he had added some aluminum fluoride to lower the melting point of cryolite, was performed on February 23, 1886. The electric current ran for several hours. When Charles cooled the melt and broke it open in Julia's presence, they found several silvery buttons of aluminum. As soon as possible, Charles took the buttons to Professor Jewett, who confirmed that they were aluminum.
If aluminum oxide breaks, at least in part, into ions in molten cryolite, then the half reactions for the electrolysis process can be written as:
The cryolite solvent that Hall found for his process has some fortunate properties in addition to being a solvent for aluminum oxide. Being ionic, cryolite is a good conductor of electricity. Although the density of solid cryolite is greater than the density of solid aluminum at room temperature, the density of molten aluminum is greater than that of molten cryolite at the electrolysis temperature. As a consequence, liquid aluminum metal collects in the bottom of the electrolysis vessel where the aluminum is protected from being reoxidized by oxygen in the atmosphere.