November 19, 1908] 



NA TURE 



81 



THE MANUFACTURE OF ARTIFICIAL 

 GRAPHITE. 



VAT^ITH the advance of electrochemistry it was found 

 that electrodes of carbon were not so satisfactory 

 as could be desired, because they disintegrated badly when 

 employed in a great many of the electrochemical processes 

 for which they were found to be practically the only sub- 

 stitute for the expensive platinum. Many attempts were 

 therefore made to convert ordinary carbon into the more 

 suitable modification — graphite — which possesses high con- 

 ductivity and resistivity. 



No very satisfactory method was devised, however, until 

 Mr. /Vcheson succeeded in obtaining a very pure form 

 accidentally when engaged in experiments upon the forma- 

 tion of silicon carbide, now known commercially as 

 carborundum. Silicon carbide can be prepared by heating 

 in the electric resistance furnace a mixture of silica and 

 carbon, when the following reaction takes place : — 



Sia-l-3C = SiC-|-2CO. 



On one occasion, by the overheating of the furnace, he 

 found that a large quantity of silicon carbide had been 

 decomposed into carbon and silicon as follows : — 



SiC = C-FSi. 



The silicon had actually been volatilised, and carbon in the 

 form of graphite remaiped behind. 



Silicon carbide is a magnificently crystalline product 

 showing a splendid iridescence, and the graphite which 

 was left behind had all the crystalline shape which the 

 carbide originally had, but it no longer possessed irides- 

 cence, and in place of being next only in hardness to the 

 diamond, was soft and friable, and had the familiar black 

 appearance of graphite. .Analysis showed it to be almost 

 pure carbon, containing no more than 0-05 per cent, of 

 impurities. 



It might naturally be supposed that the purity of the 

 final product would be directly affected by the purity of 

 the starling material. This certainly would be so if it 

 were simply a case of the high temperature of the furnace 

 changing ordinary carbon into graphite; but as the re- 

 action depends upon the formation of a carbide, which 

 then at the great heat to which it is subjected loses one 

 constituent by volatilisation, the other constituent remaining 

 behind, and, furthermore, as at the temperature at which 

 this reaction takes place all other metals are volatilised, 

 or first converted into carbides and then volatilised, leaving 

 the carbon, this is not so. In the preparation of graphite 

 it is not necessary that sand in the proportions 



SiO,-(-3C = SiC-f-2CO 



should be employed, because if the lower layers are con- 

 verted into carbide volatilisation takes place. The 

 vapourised silicon then passes through the next layer, and 

 is converted into carbide ; further volatilisation ensues, and 

 the silicon passes through another layer, and so on pro- 

 gressively, until, finally, it passes out of the furnace as 

 vapour. 



The furnace is a long, oblong brick channel with 

 electrodes at each end, which are connected together by 

 means of a core made of granulated carbon. This core is 

 surrounded by the mixture of sand and carbon, such as 

 coke, from which the graphite is to be prepared. The 

 furnace core is made of such dimensions that when the 

 current is passed the temperature will be raised sufificiently 

 high to convert the mixture surrounding it into car- 

 borundum, and then volatilise the silicon, leaving behind 

 the graphite. 



The carbon electrodes consist of twenty-five carbon rods, 

 every rod being 86 cm. long and 10 cm. square. The 

 internal part of the furnace is lined with silicon carbide, 

 which protects the fire-bricks from the enormous operating 

 heat, (lenerally, coal is employed as a raw material, being 

 first broken to the size of peas ; but as anthracite docs 

 not convey the current well, a conducting core of carbon 

 rods is run through the centre of the furnace. When 

 the furnace has been built up it is covered in with a 

 mixture of sand and coke to prevent access of air. It is 

 usually about metres long, and the anthracite layer is 



NO. zr.^-^S,, vol.. ■;()] 



50 cm. by 35 cm. square. Such a furnace requires about 

 800 kilowatts to work it. 



At the commencement of the operation the resistance is- 

 very high, therefore a high electrical pressure is required 

 at the terminal electrodes. As the temperature rises and 

 the core becomes graphitised, the resistance becomes less ; 

 consequently the voltage of the external circuit must be 

 cut down. A pressure of about 200 volts is required at 

 the commencement, but towards the end this is lowered to 

 75 volts. Shortly after the current is switched on, the 

 volatile portions of the coal are driven off and burn with 

 a characteristic yellow flame, which after a time becomes 

 less, and its place is taken by the blue-coloured flame of 

 carbon monoxide, because at this point the carbide com- 

 mences to form. As the process continues the flame again 

 becomes yellow, the carbide at this stage being decom- 

 posed. When the reduction is completed the flame is 

 cliiefly made up of burning volatilised metal, and shows 

 a very fine absorption spectra. If a cold body is intro- 

 duced into the flame it becomes covered with a felt-like ' 

 coating of silicon dioxide. 



The length of time required depends upon the purity of 

 the graphite required. For most technical purposes it may 

 contain up to 10 per cent, of ash. Further heating lowers 

 the ash content, but, of course, as more energy is re- 

 quired, it increases the cost of the finished product. 



Owing to the fact already mentioned, that much less 

 than the theoretical amount of carbide-forming material 

 need be added, Acheson finds that anthracite coal, which 

 in its natural condition contains disseminated throughout 

 its mass certain impurities, such as Fe,0,, SiO,, .AUO3, 

 &c., is particularly well adapted to produce graphite. That 

 which contains 5-78 per cent, of impurities is especially 

 suited for tliis purpose, and the graphite obtained from 

 it contains only about 0-033 per cent, of ash. Other 

 carbonaceous materials, such as brown coal, may be used,, 

 but, as a rule, the results are not so satisfactory. 



It is not necessary that the carbide should be silicon 

 carbide; other carbides, such as iron, appear to be of 

 equal value in the formation of graphite. A spectro- 

 graphic study of the flames produced in the operation 

 always shows the presence of volatilised metals. 



A particularly fine quality of graphite can be obtained 

 from the coke left at the bottom of the stills used for 

 cracking petroleum. This coke, known as petroleum coke, 

 is honeycombed with small holes produced by the escaping 

 gases, and the graphite obtained has exactly the same 

 appearance as the original product, except that in place of 

 the dull black of the coke it has the polished appearance 

 characteristic of graphite. 



For making electrodes, crucibles, motor brushes, and 

 other articles of any particular shape, it is not necessary, 

 as is the case with natural graphite, to grind up and 

 then shape the articles with some suitable binding 

 material ; but by the Acheson process the articles are first 

 made from some form of amorphous carbon, and after- 

 wards converted directly into graphite. For example, 

 lamp-black, powdered wood charcoal, or coke is mixed 

 with a metal, its oxide or salt, which is capable of form- 

 ing a carbide ; the mixture is then moistened with water 

 containing a little sugar — molasses or other binding 

 material — made into a paste, and formed into any desired 

 shape. It is then placed in the electric furnace, embedded 

 in broken carbon, and covered over with amorphous 

 carborundum to prevent loss of heat. The dimensions of 

 the furnace are so arranged that when the current is 

 passed the temperature is raised sufficiently high to 

 graphitise a portion of the carbon ; the furnace then be- 

 comes more conducting, and the pressure at the terminals 

 ran be lowered. As an example of the proportions of 

 carbon and oxide used, the following has been found satis- 

 factory for the formation of motor brushes : — ninety-seven 

 parts finely powdered wood charcoal and three parts of 

 iron oxide mixed into a paste and formed into the desired 

 shape. If the electrodes are separated about :; metres, the 

 space between being filled with the articles to be 

 graphitised and packed with carbon, an initial F..M.F. of 

 150 volts causes a current of 300 amperes. .As the carbon 

 and ai tides become graphitised the F..M.F. drops, until 

 with a pressure of too volts a current of 7000 amperes 

 is passing. 



