THE U.S. ALUMINUM INDUSTRY OWES its existence to the remarkable work Charles Martin Hall did a century ago when he discovered a way to make pure aluminum metal. Norman C. Craig of Oberlin College has recently retraced Hall's steps. In describing Craig's adventure I rely on manuscripts by him and on historical research by Emily Nunn and William Bigglestone.

Hall made his discovery at the age of 22, eight months after he was graduated from Oberlin. His laboratory was in a woodshed behind his parents' house, his primitive equipment included several items he had made himself.

Although he had pored over chemistry books since he was quite young, he had taken only two quarters of chemistry at college. His teacher there was Frank F. Jewett.

Jewett lectured on the properties of aluminum, displaying a lump of the metal he had brought back from Europe. The chemical methods then employed to make the metal were so costly that aluminum sold for about the same price as silver. Jewett told his students that the person who found a method for producing aluminum commercially would benefit the world and make himself rich. Hall, who had already studied aluminum metallurgy, turned to another student and said, "I'm going for that metal.''

At the time the method of making aluminum metal had not advanced much beyond the work done in 1825 by Hans Christian Oersted, the Danish scientist who is remembered for his pioneering research on electricity. Oersted heated a mixture of dilute potassium amalgam (potassium metal dissolved in mercury) and aluminum chloride. After removing the mercury by distillation, he found a small, gray lump of material that he supposed was aluminum metal. In 1827 the German chemist Friedrich Wohler undertook to repeat the experiment. Failing to do so, he substituted potassium metal for the potassium amalgam. By means of this substitution the experiment yielded aluminum only as a black powder. Not until 1847 did Wohler succeed in obtaining a sample of aluminum that was large enough to reveal the characteristics of the metal.

In 1854 the French chemist Henri Etienne Sainte-Claire Deville improved on Wohler's procedure by substituting sodium metal for potassium metal. A sample of aluminum produced by Sainte-Claire Deville's technique was exhibited at the 1855 Paris Exposition. The metal was demonstrably strong, light in weight, durable and resistant to corrosion. Among those who were impressed was Napoleon III, who envisioned outfitting his army with aluminum equipment for the impending war with Prussia. Commissioned by Napoleon, Sainte-Claire Deville improved his techniques, but aluminum still cost too much for wide use. From 1860 to 1880 the worldwide production of aluminum was about 1.5 tons per year.

By the 1880's the world knew of aluminum's industrial possibilities. Hall started his work against this background. Initially he tried, as other experimenters had, to reduce aluminum from aluminum oxide (alumina) by purely chemical means. He and his older sister, Julia, prepared the aluminum oxide from household alum and washing soda. (Julia helped in many other ways, drawing on her college background in chemistry.)

Hall also attempted to reduce the cost of producing aluminum with the Sainte-Claire Deville method by seeking an inexpensive way of preparing anhydrous aluminum chloride. In addition he experimented with the reaction between sodium metal and the sodium aluminum fluoride called cryolite (AlF₃.3NaF), trying to find a new and inexpensive chemical reduction method. None of the experiments showed promise.

While Hall was still in college he and Jewett had begun to investigate whether or not an electric current might make pure aluminum from an aluminum salt dissolved in water. The dissolved salt yields positive ions of aluminum. Hall and Jewett hoped that when current passed through the solution, electrons in the current would combine with the aluminum ions, creating aluminum metal.

These electrolysis experiments required several amperes of direct current. It was difficult to produce electricity in a small town such as Oberlin, and so Hall and Jewett resorted to a classical electrochemical cell invented earlier in the l9th century by Robert Wilhelm Bunsen, whose gas-burner design is still found in laboratories. A battery, sometimes known as a Bunsen-Grove battery, consists of one or more cells. Each cell centers on a porous ceramic cup holding concentrated nitric acid. A carbon-rod electrode is inserted in the acid. The cup rests in a container filled with a dilute solution of sulfuric acid (one part acid to 10 parts water). The electrode in the sulfuric acid is a sheet of zinc that is curved around the container until it almost forms a cylinder. The battery produces about 1.9 volts across the terminals, with the carbon rod acting as the positive electrode.

At the anode the zinc is converted into Zn²⁺ (a positive ion) with a release of electrons. At the cathode electrons combine with H⁺ and NO₃^- to yield nitrogen dioxide gas and water. Hall's battery employed seven of the cells. Depending on how he connected them, they generated between five and 10 amperes of current.

The initial electrolysis experiments were carried out with aluminum fluoride dissolved in water. To Hall and Jewett's disappointment the products of the reaction were hydrogen gas and aluminum hydroxide at the cathode. These experiments were done in Jewett's laboratory during Hall's senior year. After his graduation in 1885 Hall moved most of his research to the woodshed.

Since water solutions of aluminum salts had proved unpromising, Hall began to consider dissolving aluminum oxide in fused salts: salts heated to their melting point. Most fused salts are liquids similar to water in viscosity and thermal conductivity. Moreover, they conduct electricity well. (The idea of dissolving aluminum oxide in anything was bold at the time because the material was regarded as inert. For this reason it was commonly included in firebricks.)

The coal-burning furnace Hall had used in earlier experiments could not generate and sustain the high temperature needed for the work with aluminum. Hall therefore set up a gasoline burner to heat the interior of an iron tube that he lined with clay. Although the new furnace could melt some fluoride salts such as potassium and sodium fluorides, it could not melt others such as calcium fluoride, aluminum fluoride and magnesium fluoride, each of which liquefies only at a temperature of at least 1,260 degrees Celsius. Unfortunately the pure fluoride salts that did melt in the furnace dissolved little aluminum oxide.

Hall was an avid reader of Scientific American, which was then a weekly periodical. Craig believes Hall's work in electrolysis was spurred by a short item in the issue for October 24, 1885, reporting the success of a Mr. Graetzel in producing pure magnesium from a fused chloride salt through an electrolytic process. Hall later acknowledged that the process he discovered was similar to Graetzel's.

Hall still had to deal with the fact that his furnace could not produce temperatures high enough to melt some of the pure salts that might dissolve appreciable amounts of aluminum oxide. He decided to experiment with cryolite, a mixture of sodium fluoride salt and aluminum fluoride salt. Salt mixtures often have lower melting points than the pure salts from which they are made. Hall synthesized a sample of cryolite and demonstrated that its melting point was within the capability of his furnace. In later tests he added aluminum fluoride to reduce the melting point even more. When he added aluminum oxide to the hot liquid, he found that it easily dissolved. By February 10, 1886, the experiments were completed.

Six days later Hall did another crucial experiment. Could he produce pure aluminum by sending an electric current through the liquid salt in which aluminum oxide had been dissolved? He inserted two graphite rods into the hot solution, which was held in a clay crucible. Connecting one rod to the low-potential end of the Bunsen-Grove battery and the other rod to the high-potential end, he allowed the current to flow for about two hours. Then he poured the solution into a frying pan. When the material had cooled and solidified, he broke it apart. The experiment had failed. The deposit on the cathode was gray, not shiny as he knew aluminum should be.

Hall repeated the experiment with the same results. Eventually he realized the gray deposit probably came from the silica in the clay of the crucible. The silica dissolved in the hot solution, and then silicon was reduced to its elemental form by the electric current. Hall switched to a graphite crucible. The replacement was no mean feat, because large graphite rods were hard to obtain in the 1880's. Craig believes Hall got the graphite from nearby Cleveland, where the Brush Electric and Power Company had a supply of the material for its work in the new industry of arc lighting and the manufacture of dynamos.

On February 23, with the graphite crucible in place, Hall repeated the experiment. After the current had flowed for two hours, he disconnected the battery, poured out the solution and let it cool and solidify. Then he broke apart the solid material. Inside were several shiny nuggets of pure aluminum.

Hall was unaware that he had been in a race with Paul L. T. Heroult, a Frenchman who discovered a similar process and filed for a French patent in April. Today the basic electrolytic process for the production of aluminum metal goes by the name of the Hall-Heroult process.

When aluminum oxide dissolves in the hot cryolite melt, it probably dissociates into aluminum ions (Al³⁺) and oxygen ions (0^2- ), At the anode the oxygen ions combine with carbon in the electrode to become carbon dioxide gas, which then bubbles from the liquid. The conversion releases electrons that are pulled through the electrode to the high-potential end of the battery. Electrons are expelled from the low-potential end; they flow to the other electrode in the hot melt, where they combine with the aluminum ions to form liquid aluminum. As the aluminum accumulates it combines into globules that sink to the bottom of the melt owing to the lower density of the surrounding electrolyte. The globules cool and solidify, forming small nuggets of aluminum. Some of the nuggets obtained by Hall are currently on display by Alcoa (the Aluminum Company of America). The company refers to the nuggets as "the Crown Jewels" because of their brilliance and historical significance.

The maximum amount of aluminum reduced in the experiment can be calculated from the amount of current flowing through the liquid. Three electrons are needed to reduce one Al3+ ion to a neutral Al atom. Three moles of electrons are needed to reduce one mole of aluminum. A mole of electrons has a charge equal to the product of 6.022 X 10²³ (Avogadro's number) and 1.6 X 10^-19 coulombs (the charge of each electron). Thus a mole of electrons has a charge of 9.64 X 10⁴ coulombs. Suppose the current through the liquid is five amperes, that is, five coulombs per second. If the current l flows for two hours, the total charge 9 sent through the liquid is 3.6 X 10⁴ coulombs, or .37 mole of electrons. The current reduces a total of .12 mole of aluminum. One mole of aluminum

has a mass of 26.98 grams. Hence the current produces about 3.4 grams of aluminum. Aluminum's density is 2.698 grams per cubic centimeter. If all the aluminum were collected into one spherical nugget, the diameter of he nugget would be 1.3 centimeters. n practice the amount of aluminum metal produced is about half the theoretical maximum because of inefficiencies in the process.

As Hall noted in a letter to his brother on the day of the successful experiment, the cryolite is mostly unchanged by the passage of current and so the process can be continued. As the initial amount of aluminum is reduced to a pure metal and taken out of solution, more aluminum oxide can be dissolved in the fused salt. Pure aluminum continues to form as long as the current flows and the anode (which is being consumed) is pushed farther into the liquid.

Craig re-created Hall's experiment in an electric pot furnace capable of reaching 1,000 degrees C. The ingredients are melted in a number 0000 graphite crucible of the "plumbago" type (volume 118 milliliters). This crucible in turn sits in a nickel crucible that protects against leaks in case the graphite cracks while it is being heated. The nickel crucible rests on a strip of soft iron three millimeters thick and 19 millimeters wide. The strip is bent so that it extends under the crucible and also up through the top opening of the furnace. The strip serves as the connection to the low-potential end of the source of current. The strip, the nickel crucible, the graphite crucible and the hot liquid form part of the electrical pathway.

The electrical connection Craig makes to the high-potential end of the current source is through a graphite rod immersed in the hot liquid. The rod is eight millimeters in diameter and 30.5 centimeters long. The rod and the iron strip are connected to a source of direct current by large alligator clips. The rod is held in place by an insulated clamp. The temperature of the furnace is monitored with a chromel-alumel thermocouple that is encased in a ceramic sheath and extends to the bottom of the furnace. A chemical hood set in place over the equipment draws off the fluoride fumes generated by the experiment.

Craig fills the crucible with a mixture of 130 grams of sodium fluoraluminate (cryolite) and 40 grams of aluminum fluoride. When the mixture melts at 830 degrees C., he adds three grams of aluminum oxide in the form of a fine powder. He then inserts the graphite rod and connects the current source. The difference in voltage between the rod and the iron strip is about five volts and the current is about 10 amperes.

A ceramic lid covers most of the top opening of the furnace to reduce the heat loss. The graphite rod extends through the remaining opening, which Craig packs with quartz wool. The furnace is operated at a temperature of about 930 degrees C. The current is kept on for four hours, during which time another two grams of aluminum oxide is added to replenish the amount in solution. As the graphite rod is consumed, it is pushed farther into the liquid. At least once during the experiment the rod is so nearly consumed that it must be replaced.

After the furnace has cooled, the graphite crucible is taken out of it and cracked open with a hammer. At the bottom of the crucible lies a nugget of aluminum roughly one centimeter in diameter. The procedure usually yields several smaller nuggets.

Craig suggests that several changes can be made to reduce the cost of the experiment. The crucible could be heated with a butane torch in a furnace of the Hall type made from a ceramic drain tile. The electrical source can be an automobile storage battery to which an adjustable power resistor is added to decrease the voltage. The nickel crucible could be replaced with an iron crucible or possibly a clay flowerpot. The conducting strip of iron could be replaced with a second graphite rod immersed in the liquid in the same way as Hall did it. The thermocouple is not necessary because the melting of the materials in the crucible reveals the temperature. You may find other changes that can be made. Be careful, however, to do the experiment protected by a chemical hood or some other means of good ventilation. Also take precautions against spilling the hot liquid. In addition you must guard carefully against electric shock if the electrical source you are employing is at high voltage.

You might also enjoy investigating a Bunsen-Grove battery, but be extremely careful with the acids and work in a well-ventilated area. Craig constructed one of these batteries in a 400-milliliter beaker. A sheet of zinc served as one of the electrodes. The sheet curved along the wall of the beaker but did not form a complete circle Craig measured an output of four amperes from the battery. If you repeat Hall's experiment with such a battery, be sure to have enough zinc on hand. About one pound of zinc is consumed for each ounce of aluminum that is produced.

Bibliography

THE IMMORTAL WOODSHED: THE STORY OF THE INVENTOR WHO BROUGHT ALUMINUM TO AMERICA. Junius David Edwards. Dodd, Mead, 1955.

CHARLES MARTIN HALL-THE YOUNG MAN, HIS MENTOR, AND HIS METAL. N. C. Craig in Journal of Chemical Education, Vol. 63, No. 7; July, 1986.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.