Chapter 24

The precipitation of the copper from the solution, in which it is present as sulphate, or as cuprous and cupric chlorides, is generally effected by metallic iron. Either wrought, pig, iron sponge or iron bars are employed, and it is important to notice that the form in which the copper is precipitated, and also the time taken for the separation, largely depend upon the condition in which the iron is applied. Spongy iron acts most rapidly, and after this follow iron turnings and then sheet clippings. Other precipitants such as sulphuretted hydrogen and solutions of sulphides, which precipitate the copper as sulphides, and milk of lime, which gives copper oxides, have not met with commercial success. When using iron as the precipitant, it is desirable that the solution should be as neutral as possible, and the quantity of ferric salts present should be reduced to a minimum; otherwise, a certain amount of iron would be used up by the free acid and in reducing the ferric salts. Ores in which the copper is present as sulphate are directly lixiviated and treated with iron. Mine waters generally contain the copper in this form, and it is extracted by conducting the waters along troughs fitted with iron gratings.

The wet extraction of metallic copper from ores in which it occurs as the sulphide, may be considered to involve the following operations: (1) conversion of the copper into a soluble form, (2) dissolving out the soluble copper salt, (3) the precipitation of the copper. Copper sulphide may be converted either into the sulphate, which is soluble in water; the oxide, soluble in sulphuric or hydrochloric acid; cupric chloride, soluble in water; or cuprous chloride, which is soluble in solutions of metallic chlorides.

The conversion into sulphate is generally effected by the oxidizing processes of weathering, calcination, heating with iron nitrate or ferric sulphate. It may also be accomplished by calcination with ferrous sulphate, or other easily decomposable sulphates, such as aluminium sulphate. Weathering is a very slow, and, therefore, an expensive process; moreover, the entire conversion is only accomplished after a number of years. Calcination is only advisable for ores which contain relatively much iron pyrites and little copper pyrites. Also, however slowly the calcination may be conducted, there is always more or less copper sulphide left unchanged, and some copper oxide formed. Calcination with ferrous sulphate converts all the copper sulphide into sulphate. Heap roasting has been successfully employed at Agordo, in the Venetian Alps, and at Majdanpek in Servia. Josef Perino's process, which consists in heating the ore with iron nitrate to 50°-150° C., is said to possess several advantages, but it has not been applied commercially. Ferric sulphate is only used as an auxiliary to the weathering process and in an electrolytic process.

The conversion of the sulphide into oxide is adopted where the Douglas-Hunt process is employed, or where hydrochloric or sulphuric acids are cheap. The calcination is effected in reverberatory furnaces, or in muffle furnaces, if the sulphur is to be recovered. Heap, stall or shaft furnace roasting is not very satisfactory, as it is very difficult to transform all the sulphide into oxide.

The conversion of copper sulphide into the chlorides may be accomplished by calcining with common salt, or by treating the ores with ferrous chloride and hydrochloric acid or with ferric chloride. The dry way is best; the wet way is only employed when fuel is very dear, or when it is absolutely necessary that no noxious vapours should escape into the atmosphere. The dry method consists in an oxidizing roasting of the ores, and a subsequent chloridizing roasting with either common salt or _Abraumsalz_ in reverberatory or muffle furnaces. The bulk of the copper is thus transformed into cupric chloride, little cuprous chloride being obtained. This method had been long proposed by William Longmaid, Max Schaffner, Becchi and Haupt, but was only introduced into England by the labours of William Henderson, J. A. Phillips and others. The wet method is employed at Rio Tinto, the particular variant being known as the "Dötsch" process. This consists in stacking the broken ore in heaps and adding a mixture of sodium sulphate and ferric chloride in the proportions necessary for the entire conversion of the iron into ferric sulphate. The heaps are moistened with ferric chloride solution, and the reaction is maintained by the liquid percolating through the heap. The liquid is run off at the base of the heaps into the precipitating tanks, where the copper is thrown down by means of metallic iron. The ferrous chloride formed at the same time is converted into ferric chloride which can be used to moisten the heaps. This conversion is effected by allowing the ferrous chloride liquors slowly to descend a tower, filled with pieces of wood, coke or quartz, where it meets an ascending current of chlorine.

The sulphate, oxide or chlorides, which are obtained from the sulphuretted ores, are lixiviated and the metal precipitated in the same manner as we have previously described.

The metal so obtained is known as "cement" copper. If it contains more than 55% of copper it is directly refined, while if it contains a lower percentage it is smelted with matte or calcined copper pyrites. The chief impurities are basic salts of iron, free iron, graphite, and sometimes silica, antimony and iron arsenates. Washing removes some of these impurities, but some copper always passes into the slimes. If much carbonaceous matter be present (and this is generally so when iron sponge is used as the precipitant) the crude product is heated to redness in the air; this burns out the carbon, and, at the same time, oxidizes a little of the copper, which must be subsequently reduced. A similar operation is conducted when arsenic is present; basic-lined reverberatory furnaces have been used for the same purpose.

_Electrolytic Refining._--The principles have long been known on which is based the electrolytic separation of copper from the certain elements which generally accompany it, whether these, like silver and gold, are valuable, or, like arsenic, antimony, bismuth, selenium and tellurium, are merely impurities. But it was not until the dynamo was improved as a machine for generating large quantities of electricity at a very low cost that the electrolysis of copper could be practised on a commercial scale. To-day, by reason of other uses to which electricity is applied, electrically deposited copper of high conductivity is in ever-increasing demand, and commands a higher price than copper refined by fusion. This increase in value permits of copper with not over £2 or $10 worth of the precious metals being profitably subjected to electrolytic treatment. Thus many million ounces of silver and a great deal of gold are recovered which formerly were lost.

The earliest serious attempt to refine copper industrially was made by G. R. Elkington, whose first patent is dated 1865. He cast crude copper, as obtained from the ore, into plates which were used as anodes, sheets of electro-deposited copper forming the cathodes. Six anodes were suspended, alternately with four cathodes, in a saturated solution of copper sulphate in a cylindrical fire-clay trough, all the anodes being connected in one parallel group, and all the cathodes in another. A hundred or more jars were coupled in series, the cathodes of one to the anodes of the next, and were so arranged that with the aid of side-pipes with leaden connexions and india-rubber joints the electrolyte could, once daily, be made to circulate through them all from the top of one jar to the bottom of the next. The current from a Wilde's dynamo was passed, apparently with a current density of 5 or 6 amperes per sq. ft., until the anodes were too crippled for further use. The cathodes, when thick enough, were either cast and rolled or sent into the market direct. Silver and other insoluble impurities collected at the bottom of the trough up to the level of the lower side-tube, and were then run off through a plug in the bottom into settling tanks, from which they were removed for metallurgical treatment. The electrolyte was used until the accumulation of iron in it was too great, but was mixed from time to time with a little water acidulated by sulphuric acid. This process is of historic interest, and in principle it is identical with that now used. The modifications introduced have been chiefly in details, in order to economize materials and labour, to ensure purity of product, and to increase the rate of deposition.

The chemistry of the process has been studied by Martin Kiliani (_Berg- und Hüttenmännische Zeitung_, 1885, p. 249), who found that, using the (low) current-density of 1.8 ampere per sq. ft. of cathode, and an electrolyte containing 1½ lb. of copper sulphate and ½ lb. of sulphuric acid per gallon, all the gold, platinum and silver present in the crude copper anode remain as metals, undissolved, in the anode slime or mud, and all the lead remains there as sulphate, formed by the action of the sulphuric acid (or SO4 ions); he found also that arsenic forms arsenious oxide, which dissolves until the solution is saturated, and then remains in the slime, from which on long standing it gradually dissolves, after conversion by secondary reactions into arsenic oxide; antimony forms a basic sulphate which in part dissolves; bismuth partly dissolves and partly remains, but the dissolved portion tends slowly to separate out as a basic salt which becomes added to the slime; cuprous oxide, sulphide and selenides remain in the slime, and very slowly pass into solution by simple chemical action; tin partly dissolves (but in part separates again as basic salt) and partly remains as basic sulphate and stannic oxide; zinc, iron, nickel and cobalt pass into solution--more readily indeed than does the copper. Of the metals which dissolve, none (except bismuth, which is rarely present in any quantity) deposits at the anode so long as the solution retains its proper proportion of copper and acid, and the current-density is not too great. Neutral solutions are to be avoided because in them silver dissolves from the anode and, being more electro-negative than copper, is deposited at the cathode, while antimony and arsenic are also deposited, imparting a dark colour to the copper. Electrolytic copper should contain at least 99.92% of metallic copper, the balance consisting mainly of oxygen with not more than 0.01% in all of lead, arsenic, antimony, bismuth and silver. Such a degree of purity is, however, unattainable unless the conditions of electrolysis are rigidly adhered to. It should be observed that the free acid is gradually neutralized, partly by chemical action on certain constituents of the slime, partly by local action between different metals of the anode, both of which effect solution independently of the current, and partly by the peroxidation (or aëration) of ferrous sulphate formed from the iron in the anode. At the same time there is a gradual substitution of other metals for copper in the solution, because although copper _plus_ other (more electro-positive) metals are constantly dissolving at the anode, only copper is deposited at the cathode. Hence the composition and acidity of the solution, on which so much depends, must be constantly watched.

The dependence of the mechanical qualities of the copper upon the current-density employed is well known. A very weak current gives a pale and brittle deposit, but as the current-density is increased up to a certain point, the properties of the metal improve; beyond this point they deteriorate, the colour becoming darker and the deposit less coherent, until at last it is dark brown and spongy or pulverulent. The presence of even a small proportion of hydrochloric acid imparts a brown tint to the deposit. Baron H. v. Hübl (_Mittheil. des k. k. militär-geograph. Inst._, 1886, vol. vi. p. 51) has found that with neutral solutions a 5% solution of copper sulphate gave no good result, while with a 20% solution the best deposit was obtained with a current-density of 28 amperes per sq. ft.; with solutions containing 2% of sulphuric acid, the 5% solution gave good deposits with current-densities of 4 to 7.5 amperes, and the 20% solution with 11.5 to 37 amperes, per sq. ft. The maximum current-densities for a _pure_ acid solution at rest were: for 15% pure copper sulphate solutions 14 to 21 amperes, and for 20% solutions 18.5 to 28 amperes, per sq. ft.; but when the solutions were kept in gentle motion these maxima could be increased to 21-28 and 28-37 amperes per sq. ft. respectively. The necessity for adjusting the current-density to the composition and treatment of the electrolyte is thus apparent. The advantage of keeping the solution in motion is due partly to the renewal of solution thus effected in the neighbourhood of the electrodes, and partly to the neutralization of the tendency of liquids undergoing electrolysis to separate into layers, due to the different specific gravities of the solutions flowing from the opposing electrodes. Such an irregular distribution of the bath, with strong copper sulphate solution from the anode at the bottom and acid solution from the cathode at the top, not only alters the conductivity in different strata and so causes irregular current-distribution, but may lead to the current-density in the upper layers being too great for the proportion of copper there present. Irregular and defective deposits are therefore obtained. Provision for circulation of solution is made in the systems of copper-refining now in use. Henry Wilde, in 1875, in depositing copper on iron printing-rollers, recognized this principle and rotated the rollers during electrolysis, thereby renewing the surfaces of metal and liquid in mutual contact, and imparting sufficient motion to the solution to prevent stratification; as an alternative he imparted motion to the electrolyte by means of propeller blades. Other workers have followed more or less on the same lines; reference may be made to the patents of F. E. and A. S. Elmore, who sought to improve the character of the deposit by burnishing during electrolysis, of E. Dumoulin, and Sherard Cowper-Coles (_Engineering Review_, 1905, vol. xiii. p. 392), who prefers to rotate the cathode at a speed that maintains a peripheral velocity of at least 1000 ft. per minute. Certain other inventors have applied the same principle in a different way. H. Thofehrn in America and J. C. Graham in England have patented processes by which jets of the electrolyte are caused to impinge with considerable force upon the surface of the cathode, so that the renewal of the liquid at this point takes place very rapidly, and current-densities per sq. ft. of 50 to 100 amperes are recommended by the former, and of 300 amperes by the latter. Graham has described experiments in this direction, using a jet of electrolyte forced (beneath the surface of the bath) through a hole in the anode upon the surface of the cathode. Whilst the jet was playing, a good deposit was formed with so high a current-density as 280 amperes per sq. ft., but if the jet was checked, the deposit (now in a still liquid) was instantaneously ruined. When two or more jets were used side by side the deposit was good opposite the centre of each, but bad at the point where two currents met, because the rate of flow was reduced. By introducing perforated shields of ebonite between the electrodes, so that the full current-density was only attained at the centres of the jets, these ill effects could be prevented. One of the chief troubles met with was the formation of arborescent growths around the edges of the cathode, due to the greater current-density in this region; this, however, was also obviated by the use of screens. By means of a very brisk rotation of cathode, combined with a rapid current of electrolyte, J. W. Swan has succeeded in depositing excellent copper at current-densities exceeding 1000 amperes per sq. ft. The methods by which such results are to be obtained cannot, however, as yet be practised economically on a working scale; one great difficulty in applying them to the refining of metals is that the jets of liquid would be liable to carry with them articles of anode mud, and Swan has shown that the presence of solid particles in the electrolyte is one of the most fruitful causes of the well-known nodular growths on electro-deposited copper. Experiments on a working scale with one of the jet processes in America have, it is reported, been given up after a full trial.

In copper-refining practice, the current-density commonly ranges from 7.5 to 12 or 15, and occasionally to 18, amperes per sq. ft. The electrical pressure required to force a current of this intensity through the solution, and to overcome a certain opposing electromotive force arising from the more electro-negative impurities of the anode, depends upon the composition of the bath and of the anodes, the distance between the electrodes, and the temperature, but under the usual working conditions averages 0.3 volt for every pair of electrodes in series. In nearly all the processes now used, the solution contains about 1½ to 2 lb. of copper sulphate and from 5 to 10 oz. of sulphuric acid per gallon of water, and the space between the electrodes is from 1½ to 2 in., whilst the total area of cathode surface in each tank may be 200 sq. ft., more or less. The anodes are usually cast copper plates about (say) 3 ft. by 2 ft. by ¾ or 1 in. The cathodes are frequently of electro-deposited copper, deposited to a thickness of about 1/32 in. on black-leaded copper plates, from which they are stripped before use. The tanks are commonly constructed of wood lined with lead, or tarred inside, and are placed in terrace fashion each a little higher than the next in series, to facilitate the flow of solution through them all from a cistern at one end to a well at the other. Gangways are left between adjoining rows of tanks, and an overhead travelling-crane facilitates the removal of the electrodes. The arrangement of the tanks depends largely upon the voltage available from the electric generator selected; commonly they are divided into groups, all the baths in each group being in series. In the huge Anaconda plant, for example, in which 150 tons of refined copper can be produced daily by the Thofehrn multiple system (not the jet system alluded to above), there are 600 tanks about 8¼ ft. by 4½ ft. by 3¼ ft. deep, arranged in three groups of 200 tanks in series. The connexions are made by copper rods, each of which, in length, is twice the width of the tank, with a bayonet-bend in the middle, and serves to support the cathodes in the one and the anodes in the next tank. Self-registering voltmeters indicate at any moment the potential difference in every tank, and therefore give notice of short circuits occurring at any part of the installation. The chief differences between the commercial systems of refining lie in the arrangement of the baths, in the disposition and manner of supporting the electrodes in each, in the method of circulating the solution, and in the current-density employed. The various systems are often classed in two groups, known respectively as the _Multiple_ and _Series_ systems, depending upon the arrangement of the electrodes in each tank. Under the multiple system anodes and cathodes are placed alternately, all the anodes in one tank being connected to one rod, and all the cathodes to another, and the potential difference between the terminals of each tank is that between a single pair of plates. Under the series system only the first anode and the last cathode are connected to the conductors; between these are suspended, isolated from one another, a number of intermediate bi-polar electrode plates of raw copper, each of these plates acting on one side as a cathode, receiving a deposit of copper, and on the other as an anode, passing into solution; the voltage between the terminals of the tank will be as many times as great as that between a single pair of plates as there are spaces between electrodes in the tank. In time the original impure copper of the plates becomes replaced by refined copper, but if the plates are initially very impure and dissolve irregularly, it may happen that much residual scrap may have to be remelted, or that some of the metal may be twice refined, thus involving a waste of energy. Moreover, the high potential difference between the terminals of the series tank introduces a greater danger of short-circuiting through scraps of metal at the bottom of the bath; for this reason, also, lead-lined vats are inadmissible, and tarred slate tanks are often used instead. A valuable comparison of the multiple and series systems has been published by E. Keller (see _The Mineral Industry_, New York, 1899, vol. vii. p. 229). G. Kroupa has calculated that the cost of refining is 8s. per ton of copper higher under the series than it is under the multiple system; but against this, it must be remembered that the new works of the Baltimore Copper Smelting and Rolling Company, which are as large as those of the Anaconda Copper Mining Company, are using the Hayden process, which is the chief representative of the several series systems. In this system rolled copper anodes are used; these, being purer than many cast anodes, having flat surfaces, and being held in place by guides, dissolve with great regularity and require a space of only 5/8 in. between the electrodes, so that the potential difference between each pair of plates may be reduced to 0.15-0.2 volt.

J. A. W. Borchers, in Germany, and A. E. Schneider and O. Szontagh, in America, have introduced a method of circulating the solution in each vat by forcing air into a vertical pipe communicating between the bottom and top of a tank, with the result that the bubbling of the air upward aspirates solution through the vertical pipe from below, at the same time aërating it, and causing it to overflow into the top of the tank. Obviously this slow circulation has but little effect on the rate at which the copper may be deposited. The electrolyte, when too impure for further use, is commonly recrystallized, or electrolysed with insoluble anodes to recover the copper.

The yield of copper per ampere (in round numbers, 1 oz. of copper per ampere per diem) by Faraday's law is never attained in practice; and although 98% may with care be obtained, from 94 to 96% represents the more usual current-efficiency. With 100% current-efficiency and a potential difference of 0.3 volt between the electrodes, 1 lb. of copper should require about 0.154 electrical horse-power hours as the amount of energy to be expended in the tank for its production. In practice the expenditure is somewhat greater than this; in large works the gross horse-power required for the refining itself and for power and lighting in the factory may not exceed 0.19 to 0.2 (or in smaller works 0.25) horse-power hours per pound of copper refined.