Chemistry of Cyanidation of Gold & Silver

The usual reaction given for the dissolution of gold and metallic silver in cyanide solution is known as Elsner’s equation.

2Au + 4KCN + O + H20 = 2KAu(CN)2 +2KOH

Silver sulphide, the form in which silver most commonly occurs in its ores, involves a different set of reactions, which are usually expressed thus:

First stage, Ag2S + 4KCN = 2KAg(CN)2 + K2S

This, being a reversible reaction, cannot proceed far before reaching equilibrium, unless the product K2S is removed out of the sphere of action. The potassium sulphide, however, happens to be very sensitive to oxidation so that a change rapidly takes place, probably in two directions,

1. K2S + KCN + O + H2O = KCNS + 2KOH
2. 2K2S + 2O2 + H2O = K2S2O3 + 2KOH

This thiosulphate would tend later to oxidize to sulphate,

K2S2O3 + 2KOH + 2O2 = 2K2SO4 +H2O

and perhaps also more sulphocyanate would be formed, thus,

K2S2O3 + KCN = K2SO3 + KCNS.

The action of cyanide on the haloid compounds of silver, e.g., horn silver, differs from those already given in that oxygen does not appear as an indispensable auxiliary:

AgCl + 2KCN = KAg(CN)2 + KCl

How does Cyanide Leach Silver Sulphide

In connection with the reactions usually assumed to explain the action of cyanide on silver sulphide it is interesting to note that a solution of the double cyanide of potassium and silver in presence of an excess of potassium cyanide will tolerate the presence of an appreciable amount of soluble sulphide without precipitation of the silver. The reaction expressed by the equation

Ag2S + 4KCN = K2S + 2KAg(CN)2

is reversible, and its direction depends on the relative proportions of free cyanide and soluble sulphide, and the degree of their concentration.

“The reaction has been more recently studied by Berthelot who found that in dilute (N/10) solutions nearly 100 molecules of KCN were required to balance one molecule of K2S in order to retain silver in solution, instead of four molecules as the equation seems to indicate. To be exact, the equation given by Berthelot as representing the conditions of equilibrium is

96KCN + Ag2S = 2KAg(CN)2 + 92KCN + K2S

Now this proportion (96 mol. KCN to 1 mol. Ag2S) means that 96 X 65 parts of KCN are required to dissolve and hold in solution 2 X 108 parts of silver in the form of sulphide, or 28.9 to 1, if none of the sulphide is oxidized.”

The above proportions, however, would appear to be different for different concentrations of cyanide because Sharwood states in another place that

“if we dissolve a fairly large amount of Ag2S in strong solution of KCN and dilute it shortly afterward more or less of the Ag2S will be reprecipitated proving that some of the sulphide radicle remained in the solution;”

and showing also that the degree of concentration of the cyanide as well as its amount materially affects its capacity of holding silver in solution against the opposing force of Na2S. Of course if the Na2S is removed out of the sphere of action either by oxidation or precipitation as an insoluble sulphide the equilibrium is disturbed and silver sulphide continues to dissolve.

Harai R. Layng advances an interesting theory that silver in the form of the sulphide goes into solution as a sulphocyanate dissolved in cyanide solution, with the formula AgCNS.KCN. While many might not be prepared to accept without reservation his statement, rather positively made, that silver “usually enters the solutions in this form” yet the possibility is worth considering, and may throw light on some of the obscure phenomena of the process. He does not give any equation to illustrate the reaction, but the following is suggested by the present writer.

Ag2S + 4KCN+ O+ H2O = AgCNS.KCN + KAg(CN)2+ 2KOH

What is the Function of Oxygen in cyanidation

Oxygen appears to be an indispensable factor, either directly or indirectly, in the dissolution of gold and silver by cyanide solutions, except in the case of the haloid compounds of silver. Whether the action of oxygen in the dissolution of gold be a direct one, as illustrated by the Elsner equation, or an indirect one, in the sense of acting merely the part of a depolarizing agent, as maintained by Julian and Smart, will not affect the general statement that it is a necessary adjunct for the dissolution of the precious metal.

The most generally useful agent for this purpose is atmospheric oxygen, and in many instances sufficient oxygen is absorbed by the solutions in their circulation through the plant to accomplish all that is necessary. Ores that contain reducing constituents may need more oxygen than is obtained in this way, and the additional amount is most easily supplied in the case of percolation by draining dry between washes, and by the use of air lifts or centrifugal pumps, in agitation.

The addition of chemical oxidizing agents, such as potassium ferricyanide, permanganate, sodium peroxide, and ozone, has not been attended with much success, both on account of their cost, and also because of their tendency to oxidize the cyanide to cyanate (KCNO), which is useless for the purposes of the process. They have sometimes been found useful, however, for oxidizing a highly reducing pulp before adding cyanide, as in the case of slime which has stood in dams for long periods, undergoing partial oxidation with formation of ferrous compounds.

The only reagent of the oxidizing class that has attained to any importance commercially is bromocyanogen, BrCN, and that only in the raw treatment of telluride and mispickel gold ores.

Julian and Smart consider that the activity of this compound is not due to the liberation of cyanogen, though that probably occurs, but to a liberation of oxygen, in the sense of the equation.

2BrCN + KCN + 4KOH = 2KBr + 2KCN + KCNO + 2H2O + O

Cyanide Reducing Agents

Oxygen being necessary in almost every instance for the dissolution of gold and silver by cyanide, it follows that any substance that has the property of denuding the solution of its dissolved oxygen will retard or completely stop the action of the cyanide. It is a common practice to test working solutions for their “ reducing power” by acidifying and titrating with standard permanganate. Such a test, however, is misleading in its bearing on the dissolving efficiency of a solution, because it includes the reducing effect of such substances as ferrocyanides and sulphocyanates which have not the power of absorbing oxygen from solutions and therefore do not retard the dissolving effect of the cyanide on account of their reducing character, in fact, they are not reducers when considered from that standpoint. If, then, an estimate of the detrimental reducers present be desired the ferrocyanides and sulphocyanates should be determined separately and their equivalent in terms of standard permanganate deducted from the “total reducing power,” as already found.

1 cc N/10 permanganate = 0.001619 grm. KCNS
1 cc N/10 permanganate = 0.036831 grm. K4Fe (CN)6

Among harmful reducing agents may be mentioned organic matter, sulphuretted, hydrogen, sulphurous acid, ferrous sulphide (all of which may be present in old accumulated slime dams), ferrous sulphate (which in alkaline solutions becomes ferrous hydroxide), soluble sulphides (rarely found in normal working solutions), and thiosulphates.

Effect of Temperature on cyanide leaching

The rate of dissolution of metals in cyanide solution increases with rise of temperature. Julian and Smart state that in a special series of experiments made to elucidate this point the solubility of gold increased to a maximum at 85 deg. Cent. (185 deg. F.) and then slightly decreased to the boiling point. This principle has in some instances been applied in practice, especially in the treatment of silver ores, notably in the Tonopah, Nevada, district. In most cases, however, the increased consumption of cyanide due to heating the solutions in the presence of ore more than offsets the increased extraction. From 38 deg. C. (100 deg. F.) upward the decomposition of cyanide is usually very marked, but in cold climates it is often found advantageous to maintain the temperature of the working solutions at between 15.5 and 21 deg. C. (60 to 70 deg. F. ) by artificial heating.

Dissolving Effect of Zinc-potassium Cyanide

Many writers, including Feldtmann, and Julian and Smart, maintain that gold will dissolve in pure K2Zn(CN)4 in the absence of free cyanide. If this be true of gold it would probably apply also to metallic silver, but silver compounds such as argentite do not appear to be acted on in this way. A series of experiments carried out by the writer with pure silver sulphide precipitated dried and pulverized in the laboratory showed that zinc-potassium cyanide solution equivalent to 0.3% KCN had almost no dissolving effect upon it during a 24 hour agitation in a bottle open to the air, while in solutions of the double cyanide to which varied amounts of free cyanide were added the dissolution of silver was in every case closely proportional to the quantity of free cyanide present, and corresponded with the weights of silver dissolved by solutions containing similar strengths of free cyanide, but with the double cyanide absent.

Fouling of Cyanide Solutions

This expression is indefinite, and is usually used to convey the idea of gradual deterioration in the dissolving power of a mill solution during repeated contact with the ore, due to a progressive accumulation in the solution of deleterious substances contained in the pulp. One class of such substances, already alluded to, consists of reducing agents, organic or otherwise, that have the property of denuding the solutions of their dissolved oxygen, and the remedy is obviously oxidation in some form. In regard to the accumulation of base metals, there is a tendency for the activity of the solution to be retarded by the presence of any unnecessary substance, due to an increase in viscosity and the consequent resistance to the migration of the ions. It seems doubtful, however, whether there is any ulterior and more serious harm done by the presence of these metals. Copper is often said to diminish the dissolving activity of a cyanide solution to a very serious extent, but the writer has not seen conclusive evidence of this in solutions assaying up to 2 or 3 lb. of copper per ton. Julian and Smart state that “the solubility of silver is usually small when copper is present, whereas the solubility of gold is not generally affected to any thing like the same extent.” The zinc derived from precipitation seems not to affect extraction to any appreciable extent in the majority of instances, but there are cases especially when treating ores containing antimony and arsenic, where the presence of zinc lowers the extraction of the silver as much as 10% and in a less degree that of the gold. When zinc is found to have this effect, lead in solution will often act similarly, and the addition of lead salts may have to be avoided. The writer has, however, recently come across several instances where the presence of zinc from zinc precipitation was distinctly detrimental to silver extraction though the addition of litharge was beneficial.

Strength of Solution in Cyanide

When the cyanide process was first introduced a great point was made of the selective action of weak solutions, it being asserted that a weak solution had a tendency to select the gold and silver for attack in preference to the base metals. Julian and Smart’s experiments, on the contrary, went to show that the ratio of gold to pyrite dissolved in a weak solution did not differ from the ratio of that in a strong solution. However that may be, it is a fact well known in practice that the use of a solution stronger than is necessary involves an increased chemical loss of cyanide without any corresponding gain in extraction of the precious metals. The mechanical loss in residues will obviously be greater with strong solutions than with weak, so that from every point of view it is important not to use a stronger solution than is necessary to obtain the maximum extraction that will be commercially profitable. The most suitable strength for any given ore can only be found by experiment, but, in general, silver ores require much stronger solutions than gold ores. In the treatment of slime the most usual cyanide strength for all-gold ores ranges from 0.005% to 0.05%, while for silver and silver-gold ores it maybe 0.1 % to 0.15 %, in terms of KCN, and in Pachuca strengths up to 0.4% KCN are frequently used. In the leaching of sand it is usual to employ a much stronger solution than in the agitation of slime, the idea being to obtain dissolution with a small volume of strong liquor in a short period of time so as to allow as long a time as possible for the subsequent washing out of the dissolved metals. For sand leaching, whether of gold or silver ores, the strength of cyanide commonly ranges from 0.2% to 0.3% KCN.

Sources of Cyanide Loss

The Use of Lead Salts

In the treatment of silver ores the addition of a lead compound is often a material aid to extraction. This is commonly explained by the reaction K2S + PbO + H2O = PbS + 2KOH but since K2S is rarely or never found in working solutions Clennell suggests that the lead is rather to be considered as an aid in the attack on the silver, Ag2S + PbO + 4KCN = 2KAg (CN)2 + PbS + K2O. Results obtained by various metallurgists in actual practice go to show that the efficacy of any specific lead compound is usually proportional to the lead content, and that therefore litharge is more efficient than lead acetate, weight for weight.

In regard to the amount of the lead compound used a critical point is usually found where the benefit is at a maximum and beyond which it gradually declines to zero and may even become a minus quantity.

Premature Precipitation

It sometimes happens that gold and silver which have passed into solution are precipitated while still in contact with the ore and pass out of the plant with the residue. This may result in a serious lowering of recovery and consequent loss of precious metal. The most common instance of this is probably the case of graphitic ores. The power of charcoal to precipitate gold out of cyanide solution has long been known, and it appears that graphite, and perhaps certain other organic substances, possess a similar property. W. R. Feldtmann, after an extensive investigation, concludes that gold is precipitated on charcoal and graphite not as metal but as a compound, possibly a carbonyl aurocyanide, AuCN.CO(CN)2, and that this compound is soluble to the extent of about 75% in solutions of alkaline sulphides. It has been found that the graphite in such ores is very amenable to the flotation process, so that a concentrate can be made consisting of the sulphide and nearly all the graphite. This concentrate, if gold is the only valuable constituent, is readily treated by cyanide after a dead roast. The flotation tailing, if sufficiently valuable, can then be cyanided with normal results. In cases where it is not advisable or convenient to roast a graphitic concentrate F. A. Beauchamp has found that the adverse action of the graphite may in some instances be to a large extent neutralized by giving the material a series of very short treatments, say for an hour or two hours, in cyanide solution, filtering and washing between each one. This experience would tend to show that the precipitating action of the graphite is somewhat slow.

Cyanide Regeneration

Relative Dissolving Efficiency of Sodium and Potassium Cyanide