The use of chlorine as an agent for the extraction of gold from ores was first suggested by Dr. John Percy, F.R.S., at the Swansea meeting of the British Association, in a paper embodying the results of experiments carried out later. Simultaneously, Prof. C. F. Plattner, Assay Master at the Royal Freiberg Smelting Works, applied chlorine gas to the assay of the Reichenstein residues, and proposed that a similar method of treatment should be adopted on a large scale. These residues were the result of treating the Reichenstein ore with the object of extracting the arsenic. They consisted chiefly of oxides of iron and oxidised arsenical compounds, and had been roasted in the course of the process for the extraction of the arsenic. The residues had been accumulating for more than a century, and contained from 15 dwts. to 1 oz. of gold per ton; they were considered too poor to smelt, while they could not be made to yield the gold contained in them by amalgamation.
Prof. Plattner’s suggestion was followed up by investigations made by Dr. Duflos, and by Lange. Dr. Duflos compared the results obtained by treating the residues with chlorine water by percolation in a stationary vat, and by agitation in a revolving barrel; and as these results were the same, he recommended the stationary vat as being more economical. He also obtained identical results with chlorine-water and with dilute solutions of chloride of lime and hydrochloric acid mixed together. On the other hand, Lange believed that gaseous chlorine, applied to the ore in the same manner as had been used by Plattner in assaying, was a more efficient agent than a solution of chlorine in water, and it seems to have been in accordance with his advice that the first chlorination works, that at Reichenstein. The chlorinating vessels were small earthenware pots, and the precipitant employed was sulphuretted hydrogen. Plattner subsequently recommended wooden vats coated with pitch, and ferrous sulphate as a precipitant, and although these were not at first used at Reichenstein, they were adopted by G. F. Deetken, when he introduced the system into California.
The following three systems of chlorination have been developed, and are all in use in different parts of the world:
- The Plattner process, in which the gold is dissolved by means of gaseous chlorine acting on moist ore. The actual solvent is really a saturated solution of chlorine in water.
- The Barrel process, in which a supersaturated solution of chlorine in water is used, the chlorine being kept in solution by an atmosphere of chlorine of a few pounds pressure inside the barrel.
- The Vat-solution process, in which weaker, unsaturated solutions of chlorine are used.
These processes are described separately in successive chapters, in the order of their introduction on a working scale.
Plattner Chlorination Process
The following description of the original process employed at Reichenstein is an abstract of the account given in Kerl’s Huttenkunde. The material treated consisted of the residues obtained by roasting arsenical iron pyrites for the production of white arsenic, which was volatilised and condensed in brick chambers. There were forty-eight earthen chlorination pots, each holding 150 lbs. of ore. These pots were strengthened with iron hoops, and suspended on two journals, so that they could be discharged by inverting them.
The lower part of the pots was of a conical shape, and this part was filled with pebbles and sand covered by a perforated earthen plate, the function of which was to prevent the ore from mixing with the filter bed. The ore filled the cylindrical part of the pot above the earthen plate. The chlorine was generated by the action of hydrochloric and sulphuric acids on manganese dioxide in earthenware vessels, and was conveyed thence to the ore-pots through leaden pipes. The gas was introduced below the filter bed, and passed upwards through the ore for an hour; a wooden cover was then fitted on, but not luted down until chlorine had been passed for from six to seven hours longer, after which all joints were luted down with dough, and the vat left until the next day. The cover was then removed and water, at a temperature of from 64° to 77° F., poured on, and allowed to percolate through the ore and filter bed by gravity. The liquid coming from twenty-four pots was conveyed to four vats, the first one being filled with solution before the second was used, and so on; the contents of the fourth vat, being too poor for precipitation, was used over again for leaching. The leaching was stopped when 90 cubic feet of water had passed through the total charge of 3,600 lbs., this being at the rate of 312 gallons per ton of 2,000 lbs. The liquid from the first three vats was drawn off into twenty glass globes, which were heated on a sand bath so as to raise the temperature to 77° F. Sulphuretted hydrogen, obtained from fused sulphide of lead and sulphuric acid, was passed through until the saturation point was reached, when the liquid was left to settle until the next day ; after this the clear supernatant liquid was passed through sawdust filters to catch all the sulphide of gold, which might still have been in suspension. The sulphides were refined by dissolving them in acids, precipitating the metallic gold with ferrous sulphate, and melting it in clay crucibles with nitre and borax. The amount of arsenides treated daily was 3,600 lbs., containing about 5/9 oz. of gold per ton, so that only about 250 ozs. of gold were extracted yearly, and it is difficult to believe that the enterprise could have been a commercial success.
How to Use the Plattner Chlorination Process
The system ascribed to Plattner consists of the following operations:
- The concentrates or residues are subjected to a “ dead ” roast in a reverberatory furnace.
- The roasted ore is slightly damped with water and charged into wooden vats, holding from 1 ton upwards. The vats have false bottoms consisting of filter beds of gravel or of cloth. Chlorine gas, generated in another vessel, is introduced at the bottom of the vat, and rises through the ore, permeating every part of it. The vat is then closed up and left undisturbed for twenty-four hours or more, by which time all the gold is converted into soluble chloride of gold.
- The soluble salts are then washed out by water, which is allowed to flow on to the surface of the ore, and, passing through it, drains through the filter bed. When all the gold has thus been removed in solution, the tailings are thrown away.
- The solution of gold chloride is acted on by ferrous sulphate, or some other suitable reagent, by which the gold is precipitated ; the particles of the precious metal are allowed to settle, and then are collected and melted down.
This process is still used for the treatment of small quantities of roasted concentrates in a number of mills. The principal modifications which have been made in it have been introduced mainly in order to deal with larger quantities of material.
Chlorination Vats
The vats used for impregnating the ore with chlorine are usually, in California, about 7 feet in diameter, and are made of staves 3 feet long and 2 inches thick, which consist of the best split sugar-pine. They are coated with a mixture of pitch and tar to protect them from the corrosive action of the chlorine. Before being used for the first time, the vats are thoroughly soaked with water to diminish their absorptive action on the chloride of gold solution, but all wood brought in contact with this solution is nevertheless invariably impregnated with a certain amount of gold. This may be recovered after the vats, &c., are worn out, by burning them and fusing the ashes with suitable fluxes. The false bottom generally consists of quartz pebbles, the lowest layer being of the size of hazelnuts, and each successive layer consisting of finer material until, at the top, a thin layer of fine sand (passing a 20-mesh sieve, but retained on a 60-mesh sieve) is spread evenly over the surface. The thickness of the filter bed (which is not shown in the figure) is usually from 6 to 12 inches. It is supported on boards (A, Fig. 62), 1 inch thick, in which numerous ½-inch auger holes are drilled; these boards rest on wooden strips (not shown in the figure), 3 inches wide and 1 inch thick, which do not reach the edge of the vat, and so keep a clear space 1 inch deep, just above the true bottom of the vat, in which the solution can accumulate. The solution is drawn off by a leaden pipe fitted with a stopcock, preferably of stoneware ; the pipe should be level with the bottom of the vat, which may with advantage be made with a slight fall towards the outlet to prevent any liquid being left in it. Deetken states that fine sea-shells (consisting of carbonate of lime) have been used instead of quartz pebbles for the filter beds without any prejudicial result. Talcose rocks, and particularly silicates of alumina, must not be used on account of their power of absorbing the chlorine. For the same reason sulphides, magnetic iron oxide, metallic iron, fragments of wood or other organic matter, or, briefly, any substances capable of being acted on by chlorine or of reducing the chloride of gold must be carefully excluded from the filter bed.
The surface of the filter bed may be covered with boards, not fitted closely together, but made into a framework by cross-pieces and pierced with many auger holes. This cover is useful when the tailings are being cleared out, otherwise, in shovelling away the ore, the surface of the filter bed is partly removed also. Similar devices used in the cyanide process to avoid this are given in Chap. xvi. Filter cloths of canvas, burlap, or cocoa-nut fibre matting are also frequently used above the filter-bed, stretched tightly over a framework of wood which accurately fits the inside of the vat. The space between the canvas and the wall of the vat is packed with hemp or other material closely tamped down. Filter-cloths of every material, except asbestos, are soon rotted by the action of the chlorine, and their use is frequently dispensed with. Wool lasts longer than cotton.
How to Load Ore in the VATS
When the vats are ready to be charged, a layer of dry ore is spread over the false bottom, and time given for the water from the filter bed to be drawn up into this layer by capillary attraction. If attention is not paid to this point, the lowest layer of ore becomes too wet from the combined effect of the water added to it before charging-in and that absorbed from the false bottom. The result is that the passage of the chlorine through the mass is resisted, and there is a great increase in the consumption of the gas. Deetken states that the whole of the usual charge of gas may be thus consumed, not rising more than a few inches above the bottom. The greater part of the charge is damped by sprinkling with water and thorough mixing. The amount of the water added varies with the nature of the ore, but the usual amount is from 6 to 12 per cent, for roasted ores. If it is made too wet, dry ore to the required amount is mixed with it. A good rough method of ascertaining when it is of the proper degree of dampness is to compress some in the hand; balls of ore should be readily formed in this way, but should be just dry enough to crumble up again. The reason for the addition of water is that perfectly dry chlorine has scarcely any action on metallic gold at ordinary temperatures, and up to a certain point an increase in the amount of water present raises the rate of solubility of the gold. The limit of the amount of water that can be added is, however, determined by physical conditions, as the mass must be of loose porous texture in order to permit the gas to permeate readily through every portion of it. In order to promote this porous texture and uniform dampness, the ore is shovelled upon, and made to pass through, a sieve of four holes to the linear inch. This sieve may be conveniently made to slide on rollers on iron rails placed above the vat, and the ore, shaken through it, falls into the vat in a light shower. Although left undisturbed as far as possible, the charge must be levelled off with a rake occasionally.
When the vat is filled to within 6 inches of the top, the surface of the ore is made concave or saucer-shaped, higher at the sides than in the centre. The cover, usually of wood, is then lowered on to the vat by means of a chain and pulley, and the rim luted with a mixture of wet clay and sand, or, more usually in former times, with dough. These joints are kept moist during the “gassing ” process by wet rags. The gas is introduced through a lead pipe, which is shown on the left-hand side of Fig. 62, passing into the vat below the false bottom. A small hole is left in the cover through which the displaced air may escape, and the issuing gases are tested from time to time by means of a rag tied to a stick and moistened with dilute ammonia. As soon as chlorine is found to be coming off freely, the hole is plugged, but the current of gas is not stopped until after the lapse of one or two hours more, when the charge is supposed to be saturated with the gas, the total time required for the impregnation being usually from five to eight hours.
How to Generate Chlorine Gas
The chlorine is generated in air-tight vessels of lead fitted with a stirring apparatus passing through the lid and worked from the outside. The gas necessary for a 3-ton charge of roasted concentrates may be generated in a leaden vessel of 20 inches in diameter and 12 inches deep. The joints of this generator and of all pipes traversed by the gas or by liquids carrying it in solution must be made by melting the lead, no solder being used except pure lead. The “burning together” of the joints is usually effected by an air- hydrogen or coal-gas blowpipe jet. The cover of the chlorine generator is made gas-tight by a water joint, 2 inches deep, as are also the apertures in the lid for the passage of the revolving stirrer (which is usually made of hard wood) and for the delivery tube. Heat is applied by placing the generator on a sand bath standing on a perforated arch over a fireplace. The sand bath is often replaced with advantage by a water bath, as the heat required is not more than 90° F., and sudden heating causes inconveniently tumultuous generation of gas. The charge for 3 tons of ore consists of 20 to 24 pounds of rock salt, 15 to 20 pounds of manganese dioxide, containing 70 per cent, of available material, and 35 pounds of oil of vitriol of 66° B., diluted with half its weight of water. The cover is usually removed to introduce the solids and the water, but the acid is added, half a gallon at a time, through a siphon. At Heywood’s Works, California, the acid is contained in a lead vessel furnished with a glass stopcock, from which a small continuous stream is allowed to fall into the siphon.
The outlet tube is of lead, but connections in pipes are often made by short pieces of indiarubber-tubing, well greased on the inside. These resist the action of chlorine fairly well. The gas is passed through the wash-bottle shown in Fig. 63, It is usually a large glass bottle or carboy with its bottom removed, supported in a lead-lined box filled with water. The gas is made to pass through about half an inch of water. The use of the wash-bottle is partly to free the chlorine from hydrochloric acid, or other impurities with which it is contaminated, but mainly to give an indication of the rate of flow of the gas ; it is desirable that this should be as uniform as possible, as otherwise leakage is more difficult to prevent. As soon as the current of gas falls -off, fresh acid is added and the vessel stirred. The wash-bottle, as usually constructed, would be quite inadequate to free the gas completely from hydrochloric acid, whilst, even if it were approximately eliminated, some more would be speedily formed by the decomposition of water by the chlorine. 
It is, therefore, fortunate that this elimination is not absolutely necessary, in spite of the customary declaration to the contrary which generally appears in the descriptions of the process. Thus it is frequently stated that hydrochloric acid will act on any sulphides left undecomposed in the ore, generating sulphuretted hydrogen which would precipitate the gold already dissolved in the impregnation tank. This statement would not need any criticism, if it were not for the fact that it has often been made, but has apparently never been contradicted. No doubt matters might be arranged for the above reactions to take place if sufficient care were taken, but in practice they are not to be feared for the following reasons:
- If undecomposed sulphides were present in the roasted ore, they would be attacked with much greater violence by chlorine than by hydrochloric acid, and before any gold could be dissolved, the whole of the sulphides present would be converted into sulphates, according to reactions, the final effect of which is shown in the following equation:
R2S + 4Cl2 + 4H2O = R2SO4 + 8HCl
In some cases, no doubt, chlorides would be formed.
- If the sulphides were not oxidised by the chlorine, they would be almost equally efficacious with sulphuretted hydrogen in precipitating the gold, so that even in this case hydrochloric acid would do no harm.
- If an excess of chlorine is present, sulphuretted hydrogen could scarcely be said to be formed at all under any circumstances, as, if it were formed, it would be instantly decomposed by the chlorine.
- If gold were precipitated by sulphuretted hydrogen in the impregnation vat, it would be in such a finely-divided state that it would be redissolved in chlorine in a very short time, provided the gas were in excess. The only disadvantage due to the presence of much hydrochloric acid in the gas lies in the fact that certain metallic oxides (oxides of iron, copper) are much more readily soluble in the acid than in chlorine, chlorides and water being formed, and the resulting solution will be contaminated with these chlorides, so that special precautions are necessitated to prevent the bullion from becoming base.
Ore Impregnation Time by Chlorine Gas
The ore is allowed to remain impregnated with the gas for from twenty-four to forty-eight hours, the continued presence of a strong excess of gas being ascertained at intervals by removing the plug from the cover and applying the ammonia test. When, as is usually the case, there is a large excess of gas when the impregnation is at an end, it may be disposed of in one of several ways. It may be dissolved by adding water before raising the cover (the usual method of procedure), or it may be withdrawn by aspiration and discharged outside the building, or stored in a gasometer for use in a subsequent charge. If the cover is raised before getting rid of part of the excess of the gas, the atmosphere of the mill is rendered unbearable for several minutes.
The time of impregnation varies according to the size of the particles of gold, the fineness of the metal, and the temperature employed. Chlorine has a very slow action on pure gold, the rate increasing gradually with the temperature up to 100°. These results tend to show that both chlorine and bromine dissolve gold more rapidly at 50° to 60° C. than at ordinary temperatures, and that bromine is more rapid in its action than chlorine.
The fact that the rate of dissolution of gold depends on its state of division needs no demonstration. The influence of the fineness and chemical composition of the particles of gold alloy present in the ore was pointed out by Deetken and Kustel. Fine gold is acted on more slowly than that of low standard. The alloys containing base metals (copper) are dissolved very rapidly, and small quantities even of silver appear to increase the rate of solution, but if the percentage of silver amounts to 10 per cent, or over, an insoluble coating of chloride of silver is formed over the granule, and further action is checked or completely stopped.
Chemical Reactions in the Impregnation Vat
The amount of chlorine to be used depends mainly on the substances present, other than gold, by which chlorine is absorbed. A small amount is invariably converted into hydrochloric acid by the decomposition of the water present, the extent to which this reaction proceeds being increased by light and heat. If any sulphides are present they are oxidised by the chlorine in presence of water, sulphates and hydrochloric acid being formed, as follows:
4Cl2 + 4H2O = 8HCl + 2O2
CuS + 2O2 = CuSO4
These equations do not represent the whole of the reactions that take place. Some oxygen appears to be liberated, but the subject needs investigation. If considerable quantities of hydrochloric acid are thus formed, certain sulphates are converted into chlorides and sulphuric acid is set free, the reactions being influenced by the mass (i.e., concentration) of the reagents present. Protosulphates or any other protosalts present are converted almost instantaneously to persalts by the chlorine, as follows:
6FeSO4 + 3Cl2 = 2Fe2(SO4)3 + Fe2Cl6
It is obvious from these reactions that great waste of chlorine in the impregnation vat is caused by imperfect roasting, 1 per cent, of unoxidised sulphur present in pyrites, if open to the attack of the chlorine, being enough to convert 8.9 per cent, of chlorine (or about 200 lbs. per ton of ore) into hydrochloric acid. It would, therefore, be a mistake to neglect to roast the ore dead and then try to retrieve the error by increasing the allowance of chlorine. Moreover, it demonstrates the uselessness of eliminating the hydrochloric acid from the chlorine before mixing it with the ore, and expecting in that way to prevent the ill effects produced by sulphides. The fact that many sulphides are almost instantly oxidised by even very dilute solutions of chlorine has been proved by a series of laboratory experiments by the author. The oxidation of protosalts is almost as rapid, although the percentage waste of chlorine is considerably less; thus 1 per cent, of sulphur present in the ore as ferrous sulphate will convert 1.1 per cent, of chlorine (or 24.6 lbs. per ton of ore) into hydrochloric acid.
Sulphate of copper (CuSO4) does not appear to be acted on by chlorine, but, nevertheless, when much of it is present in a roasted ore, chlorination generally seems to be rendered impracticable. This is possibly due to the fact that some sulphate of iron accompanies it. Whenever sulphates of these metals are left in the roasted ore by accident or design it is necessary to remove them by a preliminary leaching with water before the chlorine is introduced. Of course if the ore is chlorinated in tubs by gas, it must be partially dried and sieved back into the tub before impregnation can be attempted.
Organic matter is also oxidised by chlorine, although much more slowly. At ordinary temperatures, pitch and tar are almost unaffected, and the fibres of matting, canvas, &c., are acted on very gradually. Pieces of decaying wood or dried leaves must not be introduced with the water into the leaching vat, and if surface water is used it should always be carefully strained before being run in. A rough analysis of the water employed will often be serviceable, as it is frequently strongly alkaline in dry countries, and may be softened with advantage.
The absorption of chlorine by metallic oxides is the most frequent cause of waste, and, in the vat process, there are usually no efforts made to prevent this. Well roasted sesquioxide of iron (Fe2O3) is scarcely attacked by chlorine, especially if the temperature attained in the furnace has been high. If any magnetic oxide (Fe3O4), however, has been formed from over-heating, or has been originally present in the ore, the absorption of chlorine is considerably greater, ferric chloride being formed and dissolved. Protoxide of iron is instantly converted into a mixture of chloride and sesquioxide of iron. Hydrochloric acid acts more rapidly than chlorine on all these oxides, but is nevertheless very slow in dissolving the well roasted sesquioxide. Metallic iron, which is sometimes accidentally introduced, is dissolved at once, by both chlorine and HCl. The oxides of copper and zinc are quickly dissolved by chlorine, and still more readily by HCl. Lime and magnesia also readily absorb chlorine, forming hypochlorites, chlorates and chlorides, but hypochlorites are decomposed by any acid which may be present.
If any appreciable quantity of oxides capable of absorbing chlorine are present, it is cheaper to dissolve them by adding dilute sulphuric acid to the ore, and then, if possible, to leach out the soluble sulphates formed, before subjecting the ore to the action of the gas.
How much Chlorine is required to leach gold
The amount of chlorine required varies greatly, both with the nature of the ore and the manner in which it is roasted. In order to roast pyrites dead, a long time in the furnace terminating at a high temperature is necessary, and the addition of salt may be desirable in order to chloridise oxides which would otherwise absorb the more expensive chlorine in the impregnation vat. These conditions in the furnace, however, may cause enormous losses by volatilisation, the endeavour to save a few pounds of chlorine in the vat causing the loss of 30 or 40 per cent, of the gold in the furnace. In ores where the percentage of copper, &c., is not large, and where, in consequence, salt need not be used in the furnace, the roasting may be finished at a high temperature without any disadvantage, and the consumption of chlorine may be thus reduced to a very low point. Butters states that at his mill in Kennel, California, where all descriptions of concentrates and pyrites were treated by the Plattner process, the average consumption of chlorine was 12 lbs. per 2,000 lbs. of ore.
Leaching the Charge
When it is judged that the impregnation has lasted long enough for all the gold to be dissolved, the excess of chlorine gas is removed, the lid is taken off, and water is added to the charge to wash out the soluble chloride of gold. The water may be added from below, and is then either allowed to overflow at the top, or is subsequently drawn off again at the bottom, the inflow being suspended. It is far more usual, however, to pour on water at the top, and let it flow out at the bottom.
The water must be added carefully, as otherwise the ore may pack unevenly, and channels may be formed through the mass, and the leaching thus rendered imperfect. Water is usually run from a tap on to a layer of gunny-sacking placed over the ore, by which it is distributed in a fairly even manner. It has been proposed to attach a coil of lead pipes, pierced with small holes, underneath the cover, and so sprinkle the water all over the ore in fine jets. In any case, water is added until it forms a layer 2 or 3 inches deep above the surface of the ore, and it is then allowed to stand until gas bubbles have ceased to rise through it, which happens in about half an hour. The stopcock below the false bottom is then opened, and the yellow- or blue-coloured solution (coloured by salts of iron, gold, and copper), which should have a strong odour of chlorine, is run slowly through a filter, consisting of a canvas bag, into a small barrel about 18 inches in diameter and 2 feet deep, the overflow of which passes by means of a launder or by rubber hose, to the precipitating tanks. Some of the ore and sand, escaping with the solution, is deposited in the canvas bag and barrel, but, if much slimes are present in the charge, either the canvas bag becomes clogged or the solution still remains turbid when it enters the precipitating vats. Water is supplied on the top of the ore as fast as it is drained away below, care being taken not to let the surface of the ore emerge from the liquid. The leaching is continued as long as any trace of gold can be detected in the issuing liquid by protosulphate of iron. As has been explained elsewhere (p. 26),. a reaction is visible in clear solutions so long as the liquid contains more than two-thirds of a pennyweight of gold per ton of water, or one part of gold in one million of water. The strongly-coloured turbid solutions usually encountered in mills, however, are not capable of yielding distinct reactions, unless they contain much larger amounts of gold than this. In particular, when large quantities of copper salts are present in the solution, their strong bluish tints mask the slight discoloration due to a precipitate of a small quantity of metallic gold, and moreover, they appear to interfere with the precipitation itself, in some cases at least preventing it from taking place. It is always advisable to filter the solution by asbestos or filter paper before testing the liquid. Filter paper may be used, since it is very slow in precipitating gold from dilute solutions, even if they are neutral, and this action is completely stopped if free chlorine is present. It is better to test the clear liquid with stannous chloride under the conditions given at p. 27, since the presence of salts of copper does not seem to interfere with the reaction in this case, and, moreover, the amount of gold present can be determined in very dilute solutions with much greater accuracy than if ferrous sulphate is used.
Since a few minutes longer time occupied in leaching is of small moment, while the extraction of a few more grains of soluble gold from a ton of ore may be of the utmost importance in the long run, it is advisable to continue to leach at any rate until the water contains less than 1 part of gold in 5,000,000 (about 3 grains per ton), a point which can easily be determined by means of stannous chloride properly manipulated. The last charges of wash-water should not be mixed with the strong solution, but stored in other vats and used again for the first washings of other charges. In this way the amount of wash-water does not become excessive, although the tailings are cleaned effectually. Re-precipitation of dissolved gold in storage vats or impregnation vats is not to be feared, so long as there is an excess of chlorine present in the liquid, and this can easily be ensured by adding a small quantity to any solutions not smelling strongly of the gas.
The amount of water used, varies according to the richness of the ore and the method of leaching adopted. It is usually about 2 tons of water to 1 ton of ore, but in most cases part of the water is used again in the next charge.
Precipitation of Gold in Chlorination Process
The precipitating vat is of the same materials as the leaching tubs, and may be from 5 to 7 feet in diameter and 3 feet deep. There is no false bottom, and the vat is often made wider at the bottom than at the top to prevent any adherence of the gold to the sides. The wood is protected by a coating of pitch or paraffin-paint, or is left without paint of any kind. The vat receives a smooth finish inside to facilitate perfect cleaning, and is set perfectly level to avoid loss of gold while the waste liquor is being drawn off. The precipitating solution of protosulphate of iron (the reagent which has been chiefly used in practice in the vat process) is usually introduced into the precipitating vat at the beginning of the filtering operations. Care must be taken not to introduce a wasteful quantity, as it is better to make up for any deficiency when the gold- solution has all been run in. This operation is conducted in such a way as to impart a circular motion to the contents of the vat, so that the solutions are mixed without hand-stirring, but the latter is often resorted to in addition, in order to make the precipitate settle better ; flat wooden staves with round handles are used for the purpose. The contents of the vat may be tested with solutions of gold chloride and of ferrous sulphate to determine whether the precipitation is complete, and the precipitant present in excess.
If lead or lime is present, dissolved in the solution, it will be precipitated as an insoluble sulphate on the addition of the ferrous sulphate, and thus render the gold-precipitate impure and less easy to treat. The amount of lead in the solution is usually small, unless hot water has been used for leaching, and most of the lead chloride is, in any case, separated by the canvas filter. The usual method of removing the lime is to add sulphuric acid to the gold solution, and to let it stand for a few hours, when calcium sulphate crystallises out, forming a crust on the sides and bottom of the vat. The liquid is then drawn off and transferred to another vat for the precipitation of the gold. Instead of this method, Nelson A. Ferry, E.M., recommends the addition of molasses to the leach, before the addition of the sulphate of iron. He dissolves 1 gallon of molasses in 30 or 40 gallons of water, and keeps it for use, determining the quantity to be added by a laboratory experiment. He states that in this way the precipitation of the lime is prevented, but a large excess of the ferrous sulphate should be avoided, and the liquid kept slightly acid. The gold then sometimes comes down flocculent at first, but soon changes to its normal condition.
The ferrous sulphate is usually prepared on the mill by dissolving iron in sulphuric acid. When precipitation is complete the liquid is allowed to remain at rest for some time, in order to allow the gold to settle to the bottom. The old practice was to leave it “overnight,” but the length of time allowed has of late years been greatly extended. Butters states that forty-eight hours is usually sufficient, but that sixty hours is better, and the determination of the extent to which the settling has progressed may be made by tapping the solution at various heights and filtering the liquid thus obtained. When a quart of liquid, drawn from a point 2 inches above the bottom of the vat, gives only a slight dark stain to a No. 7 Swedish filter paper, on being passed through it, the settling may be regarded as complete. C. H. Aaron quotes instances where, after forty-eight hours settling, as much gold remained in suspension in the liquid which was drawn off as was equivalent to 50 cents per ton of the ore treated. The conclusion may be drawn from these statements that a certain amount of gold is inevitably lost by being carried away in suspension, but with care and patience the loss may be reduced to a low percentage, and even at the present day, in spite of the introduction of many other precipitants, ferrous sulphate is probably as widely used as it has been at any time in the past.
When the waste liquid has been drawn off by a floating siphon, more ferrous sulphate and fresh solutions from the leaching vat are poured into the vat, and the process repeated until enough gold has accumulated at the bottom to warrant a clean-up. This may take place at intervals of from a fortnight to three months. The clear liquid is drawn off as closely as possible, and the slime scooped out and filtered through paper, or, by means of a press, through canvas. Finally, the vat is thoroughly cleaned out by rinsing it with water which is run off through a plug-hole, level with the bottom, into a wash-tub. The gold precipitate is then dried carefully and fused in graphite pots, with salt, sand, nitre, borax, &c., as fluxes, according to the requirements of the case. If the precipitate contains any considerable amount of impurities (such as oxides and basic salts of iron), which is usually the case, it may be treated with hydrochloric acid before fusion. The bullion produced varies from 920 to 990 fine, the alloying metals consisting chiefly of iron and lead.
Using Liquid Chlorine
At the Utica Mine, California, the Plattner process was in use. The concentrates are roasted in long reverberatory furnaces, allowed to cool, moistened with sulphuric acid and water and sieved into wooden vats. Chlorine gas derived from a drum of liquid chlorine is then introduced into the charge, and after forty-eight hours the ore is leached in the usual way. In 1910, the cost of treatment was formerly $7.80 per ton, when the chlorine was generated from manganese dioxide, salt and sulphuric acid, and only $6.90 per ton after the introduction of liquid chlorine. One drum of liquid chlorine weighs 300 lbs., holds 115 lbs. of chlorine, and takes the place of 972 lbs. of manganese dioxide, 1,080 lbs. of salt, and 2,160 lbs. of sulphuric acid. The drum is 10 inches in diameter and 5 feet long, and costs $40 at the Utica Mine. The time of introducing the gas into a charge is reduced by its use from six or eight hours to twenty minutes.