Tungsten and Tungsten Wire History
The word “Tungsten” was probably first used by A. F. Cronstedt in 1755, who applied it to the mineral subsequently known as “scheelite,” which is the natural form of calcium tungstate. C. C. Leonhard named this mineral scheelite in 1821 in recognition of the discovery made by K. W. Scheele, in 1781, that the mineral was a compound of lime with a previously unknown acid, which he called “Tungstic Acid,” a name by which it is still known. Before Scheele made his discovery, the mineral was generally regarded as containing tin. The word tungsten denotes a substance of high density and is derived from the Swedish language, “tung,” meaning heavy, and “sten,” meaning stone.
In 1783 the Spanish brothers, J. J. and F. d’Elhujar, published the results of their investigations on wolframite carried out with the Swede, T. Bergmann, while they were working in his laboratory. They showed that this mineral contained the same tungstic acid, previously found in scheelite, but combined with iron and manganese, instead of calcium. They were also the first to record the preparation of elementary tungsten, which they made by reducing tungstic oxide with charcoal, and to which they gave the name “Wolfram.” The origin of the word wolfram is obscure. Mennicke attributes it to the alchemists, who called the metal “spuma lupi,” which means wolf spume or foam. Another suggestion is that the word is of German origin from wolf, meaning a beast of prey, and rabin or ram, which has several meanings, including dirt and soot. The word may also be derived from the Swedish word “Wolf rig,” which means eating. All these meanings are assumed to be associated with the early difficulties of extracting tin from cassiterite when it was contaminated with wolframite; the two minerals are frequently found together, and the wolfram was thought to eat the tin as a wolf eats sheep. The common termination used in mineralogy, to give the name “wolframite” to the mineral, was used in 1820 by A. Breithaupt in his book, Kurze Charalderistik des Mineral Systems.
The metal is known as tungsten in some countries and as wolfram in others, including Sweden, the country of origin of the name tungsten. The chemical symbol W, which is universally used to denote tungsten, suggests that wolfram was formerly the more generally accepted name for the element. In Britain the mineral wolframite is also known as wolfram.
For many years tungsten remained one of the rare elements, and it was not until 1847, when Oxland took out a patent for the manufacture of sodium tungstate, tungstic acid, and tungsten from cassiterite (tinstone), that the element became of any industrial importance. Oxland’s second patent, taken out in 1857, described the manufacture of the iron-tungsten alloys that form the basis of modem high-speed steels. The metal itself, however, found no application until nearly fifty years later, when it was first employed in the manufacture of filaments for electric incandescent lamps. From 1878, when Swan demonstrated his eight and sixteen-candle power carbon lamps at Newcastle, search was made for a more satisfactory filament material than carbon. The early carbon lamp had an efficiency of about 1.0 lumen per watt, which was improved during the next 20 years by modifications in methods of preparing the carbon, to about 2.5 lumens per watt. A further improvement was made in 1898 to about 3.0 lumens per watt by heating the filaments electrically in an atmosphere of petroleum vapor, which caused the deposition of carbon in the pores of the filament and gave it a bright metallic appearance. At the same time A. Von Welsbach produced the first successful metal filament was by using osmium; attempts had previously been made to use platinum, but its relatively low melting point of 1774°C. prevented its successful development. Lamps using osmium filaments had an efficiency of about 6.0 lumens per watt. Since osmium is the rarest of the platinum metals it could never have been used on a large scale. Tantalum, with a melting point of 2996°C., compared with osmium, 2700°C., was extensively used as a drawn wire from 1903 to 1911, following work by Von Bolton of Siemens and Halske. Lamps with tantalum filaments had an efficiency of about 7.0 lumens per watt. Developments in the use of tungsten started about 1904, and it has been used exclusively since about 1911. The modern tungsten filament lamp used for general lighting purposes, which employs drawn wire, has an efficiency of about 12 lumens per watt, while lamps of high wattage have efficiencies up to about 22 lumens per watt. The modem fluorescent lamp, although it employs tungsten cathodes, does not depend upon tungsten for its much higher efficiency, which is of the order of 50 lumens per watt.
In 1904 the Siemens-Halske Co. tried to apply the drawing process they had developed for tantalum to the production of filaments of the more refractory metals, tungsten, thorium, etc. The brittleness and lack of ductility of tungsten prevented their attaining success by this method, although later, in 1913-1914, it was demonstrated that fused tungsten could be rolled and drawn at very high temperatures, using very small reduction steps. By striking an arc between a tungsten rod and a partially sintered tungsten pellet in a graphite crucible, coated on the inside with tungsten metal powder and containing an atmosphere of hydrogen, small pieces of fused tungsten, about 10 mm. diameter and 20-30 mm. long, were produced, which could be worked with difficulty. It was found that working properties could be improved to some extent by the addition of thorium oxide, which reduces the tendency to develop a columnar type of structure during cooling of the fused mass. This process was never used commercially. In the same year Just and Hannaman patented a process for producing tungsten filaments by mixing the finely divided metal powder with an organic binder, extruding through dies, and heating in suitable gases to remove the binder, leaving a pure tungsten filament. During 1906-7, the well known extrusion process, which was the method by which the majority of tungsten filaments were made for the next four or five years, was developed.
The process consisted in mixing very fine black tungsten powder with dextrin or starch in order to form a plastic mass, which was forced under hydraulic pressure through a fine diamond die. The thread produced in this way was sufficiently strong to be wound on cards and dried. The filament was then cut into “hairpins,” which were heated in an inert gas to a red heat to drive out moisture and the lighter hydrocarbons. Each “hairpin” was then mounted in clips, and raised to bright incandescence by the passage of an electric current, whilst being surrounded by a gas, such as hydrogen, chosen to react with the binding material, so that pure tungsten only remained. At the highest temperature the fine particles of tungsten sintered together and formed a solid homogeneous metallic filament. These filaments, although elastic, were quite brittle, but could be formed to shape at a red heat. Just and Hannaman also developed another process at the same time. This was known as the “coating” process, and showed remarkable ingenuity. A carbon filament as small as 0.02 mm. in diameter was employed as the base, and this was coated with tungsten by raising it to incandescence in an atmosphere of hydrogen and tungsten hexachloride. The coated filament was then raised to bright incandescence in hydrogen at a pressure of about 20 mm. of mercury. The carbon core dissolved in the tungsten, forming tungsten carbide, the change being so complete that the resulting filament was tubular in cross-section, no carbon remaining in the core. The filament so obtained presented a glittering white appearance and was very fragile. The next step consisted in heating the filament in hydrogen containing steam, which oxidized the carbon and left a compact filament of pure tungsten. The filaments thus obtained were similar to those made by the extrusion process, except that they were tubular in cross-section.
Many other processes for the production of tungsten filaments appeared in the following years, but the product obtained was in all cases of the same type, namely, an elastic but brittle tungsten filament. Amongst the more important may be mentioned the colloidal method of Kuzel, first developed in 1904. By this method a gelatinous pasty mass of metallic tungsten was prepared by striking an arc between tungsten electrodes under water. The material contained no binding medium, but was itself sufficiently plastic to be extruded into fine threads. On heating these to a high temperature in hydrogen by means of an electric current, the colloidal mass was converted into crystalline metal and the filaments were in all respects similar to those produced by the ordinary extrusion process. The method was largely used on the European Continent, and to some extent in the United States.
Another method successfully developed in America in 1906 was the amalgam process. Finely divided tungsten powder was mechanically mixed with twice its weight of cadmium-mercury amalgam, from which filaments were formed by extrusion. The filaments were strong and exceedingly ductile. The amalgam was subsequently removed by volatilization at a high temperature and a pure tungsten filament was obtained. A method which achieved considerable success, and was used between 1908 and 1910 by the Siemens and Halske Co., was that of mixing tungsten metal powder with 6-10 percent of nickel, as nickel oxide, pressing the powder into ingots, and sintering in hydrogen at 1575° C. The ingots were first rolled to rod of 1 mm. diameter at 350° C., and then, with fre-quent anneals at 1500-1600° C., drawn cold to wire as fine as 0.03 mm. The drawn wire was quite ductile. The nickel was removed by heating the finished filaments in vacuo at 1500° C. A full account of this process has been given by M. Pirani. Other processes were also developed, such as drawing wires from tantalum tubes packed with tungsten powder. It was not, however, until 1909 that Coolidge, in America, was successful in making ductile tungsten from the metal powder by suitable heat treatment and mechanical working.
In all previous processes some binding agent, either organic or metallic, had been employed to give the necessary plasticity, and was subsequently removed by chemical or thermal treatment. The filaments that resulted were pure tungsten as far as analysis could show, and yet the metal was in all cases completely brittle. Even these brittle filaments, however, could be bent and worked to some extent at a relatively low temperature, and even at temperatures below that at which oxidation takes place.
The problem of making ductile tungsten did not, therefore, appear to be one of purifying the material, although it was realized that pure metal was probably essential if a ductile product was to be obtained. Rather, the problem was caused by the grain structure of the tungsten itself. By using a sufficiently high temperature initially, it was found that as the metal was subjected to mechanical work its ductility increased, until finally it became so ductile that it could be rolled, or drawn into wire, at room temperature.
Although only a small percentage of the ore which comes on the market is used for the manufacture of lamp filaments and similar products, the great importance which tungsten has assumed scientifically and technically is the outcome of work directed to its production for this purpose. The knowledge gained has also been of inestimable value to workers in the newer fields of powder metallurgy, particularly in the manufacture of hard carbides. Consideration of the stages that have been passed in the development of modern processes gives some understanding of the difficulties that have been overcome.