Table of Contents
In Liddell’s Handbook of Nonferrous Metallurgy, Vol. 2, 1945, there is to be found a very complete account of the uses of chlorine as applied to the recovery of gold and silver, in Chlorination Processes & Methods. Introducing the chapter entitled “Chlorine Metallurgical Processes” Liddell says:
Chlorine as a metallurgical agent appears to have lost ground during recent years, and much of what follows is of historical rather than operating interest. However, it is a question in the author’s mind whether the great decrease in the price of chlorine and the large sources of supply that will be available after the war do not warrant reinvestigation of the applicability of the chlorine process to present-day metallurgy. The use of chlorine can be broadly classified into the table of content here to the right:
Dry Chlorination
Chlorination Process for Gold
This process was based on the fact that chlorine, in the presence of moisture, converts gold into the trichloride AuCl3 , which is soluble in water and removed by washing, the gold being then precipitated by ferrous sulphate, sulphur dioxide, hydrogen sulphide, or charcoal. Coarse gold requires long contact and should be removed by amalgamation. Pyritic ore or concentrate requires a dead roast before chlorination; thoroughly oxidized ore may be treated directly, Basic ores—containing lime and especially magnesia—absorbed much chlorine and might become heated. This was checked by roasting with a high temperature at the finish to frit the magnesia with silica. Chlorination was suggested by Percy and by Plattner independently in 1848, though Plattner apparently made the first commercial application of the process to the arsenical ores of Rechenstein, Silesia.
The Deetken or California process was carried out in comparatively small wooden vats with bottom filter of perforated boards, resting on slats and covered with coarse gravel and sand; a vat 8 ft. in diameter with 3-ft. staves would hold about 3 tons. The crushed and roasted ore was loosely charged by sifting in a moist condition to facilitate leaching, a cover was luted on with clay or dough, and gaseous chlorine generated in a lead vessel from manganese dioxide, salt, and sulphuric acid was admitted by a lead pipe to the bottom until it could be detected at a hole in the cover. In later practice, liquid chlorine was purchased in steel cylinders. After 12 to 36 hr. contact (adding more chlorine if necessary), water was turned in at the top, any chlorine escaping at the bottom pipe being led to another vat. The yellow solution was run to the precipitating vat, and the charge washed until the effluent was colorless. The residue was then shoveled out; if much silver was present, it was transferred to another vat and leached there with hyposulphite solution. Some ores, rich in silver, were first leached with hyposulphite and then chlorinated.
The barrel process, used on a large scale in Colorado, involved the rotation of the ore in barrels of wood or heavy lead-lined steel, holding 5 to 25 tons, while chlorine was generated under pressure in the mass by means of bleaching powder and sulphuric acid. Barrels were often built with an internal filter on one side, consisting of pebbles or coarse sand confined by slotted boards and a perforated lead plate. After 3 to 6 hr. water under 20 to 40 lb. pressure was admitted by a trunnion and washed the charge in 1 to 2 hr. A ton of ore would use at least 10 lb. of bleach and 15 lb. of sulphuric acid. The ore was charged dry from a hopper and discharged by sluicing through manholes in the side.
With the exception of the Malm process, chlorine processes for raw ores of the precious metals are necessarily confined to the treatment of surface ores or to clean gold and silver sulphide ores, in which the gold and silver minerals are not combined with the base-metal sulphide minerals.
Wet Chlorination
In the Plattner process which was in constant use from 1851 to about 1916 for the treatment of pyrite concentrates, the material after dead roasting was leached in water saturated with chlorine gas. In the early days the chlorine was generated by the addition of bleaching powder and sulphuric acid, but later, in the treatment of Cripple Creek ores, the chlorine was generated by electrolysis. Lead-lined chlorination barrels provided with perforated lead filters inside were used. The gold was usually precipitated from solution by ferrous sulphate or hydrogen sulphide.
Chlorination and Amalgamation
Patio Process
The patio process for silver was invented by a miner in Pachuca, Mexico, in 1557. The dry crushed ore was moistened and mixed with copper sulphate, common salt, and mercury, the working and aging of the charge being performed in yards, or patios. The copper chlorides formed reacted with the silver to form silver chloride. The silver was soluble in the brines present during the subsequent leach, was precipitated by the mercury and immediately amalgamated. The method was not well suited to ores carrying appreciable amounts of base-metal sulphides.
Pan-Amalgamation Process
The pan-amalgamation process was a development of the ancient “cazo,” or caldron, process. In Europe it was called “barrel amalgamation,” but in the United States, where it found its greatest use, it was known as the pan-amalgamation or “Washoe process” from Washoe, Nev., where it was first operated in 1861. Liddell describes the operation as follows:
The “pan” was made wholly of cast iron or with a cast-iron bottom and wooden sides. In either case the bottom was made hollow for the introduction of steam to heat the charge. Cast-iron mullers for grinding, stirring, and amalgamating the ore were attached to a vertical shaft in the center of each pan. The capacity of each pan ranged from 0.5 to 2 tons of ore and was usually about 1½ tons.
The ore was first crushed by jaw crushers and then by stamps or ball mills. If crushed wet, the excess of water was removed by settling tanks. The crushed ore from the settling tanks was then shoveled into the pans. Salt and copper sulphate were added in the ratio of 5 to 10 lb. of salt and 2½ to 5 lb. of copper sulphate per ton of ore. When the ore is free from interfering minerals, the salt has been reduced as low as 2 lb. and the copper sulphate to 1¼ lb. per ton. Water was added in sufficient quantity to make a thin mud, and steam was admitted, not only in the jacketed bottom of the pan but sometimes into the ore itself, until the charge in the pan was maintained at the boiling point. The grinding and stirring of the charge was continued for 2 or 3 hr. in which time the chemical action was completed.
Mercury, equal to 10 per cent of the weight of the ore, was then sprinkled over the ore pulp by straining through canvas or chamois, and the stirring continued for 3 hr. longer—when amalgamation was completed.
The whole charge was then washed from the pan into a settling tank provided with radial arms and agitated under a constant flow of water until the amalgam collected in the bottom of the tank and the tailings were washed away. The amalgam was then transferred to a small pan known as the cleanup pan, where it was stirred with additional mercury and washed with water until free from ore particles. The silver and gold were finally recovered by retorting the amalgam.
The chemistry of this process is the same as that of the patio process, except that the iron of the pan and mullers also acts as a reducing agent, not only for precipitating silver in metallic state from its chlorides but also for preventing the formation of any chlorides or sulphides of mercury, and in this manner avoids the chemical loss of mercury mentioned under the patio process. The ore must be siliceous or neutral in character to avoid precipitation and the loss of effective copper salts by the carbonates of lime and magnesia, although the iron of the pan has a tendency to reduce these salts to copper and hence militate against their effectiveness.
Chloridizing Roasting and Leaching
The gradual exhaustion of oxidized ore and the increase of base-metal sulphide minerals with the silver sulphide minerals, together with increased facilities for transporting fuel and supplies, led to the introduction of chloridizing roasting and the attendant leaching processes.
In chloridizing roasting, salt in amount equal to 5 to 15 per cent of the weight of the ore is added for a sulphur content of 2.5 to 3.0 per cent, which, if deficient, is made up by the addition of pyrites.
The Wedge-type multiple-hearth furnace was most frequently used, and the temperature was not allowed to exceed about 600°C.
Patern process
The Patern process makes use of the solubility of silver chloride in sodium thiosulphate solution. Following chloridizing roasting and a warm water wash, the ore is leached with Na2S2O3. 5H2O and the silver precipitated from the purified effluent with sodium sulphide.
Augustine process
The Augustine process consisted in giving a chloridizing roast to silver ores or matte and leaching with a strong solution of common salt, which dissolves silver chloride to a limited extent. The silver was then precipitated by metallic copper, and the copper by scrap iron. The cement copper was then melted and granulated for reuse.
Tainton process
The Tainton process included the scheme of chlorinating roasting of oxidized silver lead ores and leaching with spent chloride electrolyte containing free chlorine and ferric chloride, to which bleaching powder had been added. The solution was then electrolyzed in specially designed closed cells using rotary cathodes, and the silver and lead deposit collected and melted to bullion. An alternative scheme was to deposit the silver and gold preferentially by using a lower current density. The chlorine was used to generate the hypochlorite above mentioned as well as the calcium chloride needed to control the SO4 concentration of the solutions.
Holt-Dern process
The Holt-Dern process was developed at Park City, Utah, and reached commercial operation in 1941 when a plant was constructed at Silver City by the Tintic Milling Co. The process consisted of a chloridizing roast in a Holt-Dern furnace, followed by a percolation leach with a nearly saturated solution of common salt acidified with sulphuric acid. The silver was recovered by precipitation on detinned scrap iron.
A modification of the Holt-Dern process was used at the Achotla chloridizing mill of the Cia Minera de Penoles in Guerrero, Mexico. Here the mixture of oxide ore, sulphide ore, and salt (3.3 per cent of the charge) was given a chloridizing roast in 16 Holt-Dern blast roasters. The hot calcine was transferred to leaching vats and first given a water wash, which was drained to storage for subsequent precipitation of gold and silver values on scrap iron.
A weak acid wash, derived from spraying the flue gases, was next applied, and the effluent passed over scrap-iron trays. Following prolonged washing to neutral reaction, the ore was cyanided, using solutions carrying 1 kilogram per metric ton KCN and 0.3 to 0.7 kilogram CaO. The mill had a capacity of about 6500 tons, per month, and about 90 per cent of the gold and 84 per cent of the silver in the ore were recovered by the process described.
Chlorine Volatilization of Gold
Loss of gold by volatilization was recognized from time to time and was the cause of serious monetary loss in treating gold-silver ores by chloridizing roasting, but the reason for this loss seemed to be little understood.
In 1891—1893 Croasdale discovered that a commercially complete (above 90%) volatilization of gold could be obtained from Cripple Creek ores by roasting with salt. About the same time, Pohle independently obtained similar results with silver ores from Aspen, Colo. Systematic investigation of the volatilization of metals as chlorides was begun by these men in 1898 and was carried on with a large-scale experimental plant until 1903. Numerous.investigators have worked on this process since that time.
By raising the temperature of a chloridizing roast to 1050°C. it has been found that gold, silver, copper, and lead can be commercially volatilized as chlorides from their ores and the metals recovered from the fumes. By charging the ore, salt, and sulphur mixture as quickly as possible into the hot zone of the furnace, volatilization begins at about 750°C. and is completed within 30 to 60 min. The roasting atmosphere must be kept highly oxidizing. The process is continuous, and the ore is commercially devoid of value as it discharges from the furnace.
The furnace used for this process was a regular cement kiln, 100 to 125 ft. in length, fired in the usual manner.
Gold is easily volatilized, but in what form was never definitely determined. It was generally supposed that gold trichloride was formed at low temperatures, and this was decomposed into metallic gold and chlorine at temperatures below 300°C. If this is true, the metallic gold formed from the vapors is probably colloidal and is carried out of the furnace in that form with the gases. Rose states that, when gold is heated in chlorine at atmospheric pressure, trichloride of gold is formed, which volatilizes at all temperatures above 180°C. up to and beyond 1100°C. Other metallurgists thought that gold forms a double chloride with salt or other metallic chlorides and volatilizes in that form. The theory of a double chloride seems more probable.
Silver is less easily chloridized and volatilized than any of the common metals. It seems to be extremely sensitive to atmospheric conditions in the furnace and may be affected by the gangue constituents in the ore. Silver chloride melts to a thin liquid at about 451°C. before it volatilizes, which probably accounts for its sluggish volatilization. It is much more easily volatilized in the presence of other metallic chlorides, which would indicate that it volatilizes as a double chloride.
The metallic chlorides are driven from the ore in the form of vapor, and they condense as colloidal particles of fume. Cottrell precipitation is required for complete recovery. The metals are recovered from the collected fume by substitution of one metal for another in aqueous solution or by electrolysis of the fused chlorides.
Articles in C.E. and M. Rev. for Feb. 10 and Mar. 10, 1943, describe laboratory volatilization tests on two different products:
The first was a hematitic calcine residue carrying 7.2 dwt. Au per ton. Salt (5½ per cent by weight) was added, and the change nodulated and thoroughly dried. Using a rotary kiln externally heated, with the outlet gases passing over hot charcoal, a 25-hr. run at 900 to 1000°C. resulted in 86.8 per cent of the gold being volatilized and 95.9 per cent of the volatilized gold recovered on the charcoal. It was stated that sufficient gas flow must be used to remove products of combustion or any reducing gas that would tend to break up the auric chloride.
Lake View and Star flotation concentrates were used in the second case. B. H. Moore found that, by using 10 per cent salt and heating in a closed muffle at 800°C. for 40 min., 92.9 per cent of the total gold was volatilized. The author believes that a closed furnace is essential for best results because of the rapid decomposition of AuCl2 in air. Furthermore, there is a much smaller volume of gas to be handled than in an air-swept furnace and a better chance to collect the fine gold particles, which in the laboratory was done in a water column using glass beads.
Recent work at Columbia University on the rate of solution of gold leaf showed that the addition of sodium chloride to chlorine or bromine solutions accelerates the dissolution of pure gold and of (95.8 per cent) gold leaf immersed in these solutions while sulphates and fluorides have practically no effect. The conclusions reached were that the accelerating effect of chloride ion has not been recognized heretofore but that, if it is assumed that the first reaction between chlorine and gold is
2Au + Cl2 = 2AuCl
and that the rate of solution of the gold is controlled by the rate at which the insoluble aurous chloride film is removed from the surface by the reactions:
AuCl + Cl2 + Cl‾ = AuCl4
and, AuCl + Cl¯ = AUCl2
It was shown that aurous chloride is insoluble in water but is soluble in sodium chloride. The aurous chloride is later oxidized to the auric state, but in dilute chlorine solutions this is not the rate-controlling reaction.