Art. LXXIII.—On the Action of Potassium-Cyanide Solution upon Gold.
[Read before the Auckland Institute, 24th February, 1896.]
In the Journal of the Chemical Society (Trans., 1893, pp. 724–738, and Trans., 1895, pp. 199–212) two papers on the action of cyanide upon gold have been published. These papers contain an explanation of the peculiar action of dilute solutions of potassium cyanide, and a record of a number of experiments which, I think, prove conclusively that this explanation is correct and sufficient. So far as I am aware, the statements contained in these papers have passed unchallenged in Europe. In New Zealand, however, Mr. Skey has published* some experiments from which he concluded that no satisfactory explanation had been given of the action of cyanide on gold. As such a statement is likely to mislead those who have not had an opportunity of reading the original papers on the subject, I shall briefly outline my work already referred to, and shall then consider Mr. Skey's views.
The knowledge of the subject when I began my researches was as follows: Gold was known to be soluble in potassiumcyanide solutions, but the nature of the action was disputed. Most of the text-books referred back to Elsner (J. Pr. Chem., 37, 333) and credited him with the explanation embodied in the following equation :—
4Au + 8KCN + O2 + 2OH2 = 4 AuCNKCN + 4KOH.
On the other hand, Macarthur (patentee of the cyanide process), in a paper read before the Society of Chemical Industry, (Journal, 1890, p. 270), called in question the necessity of Oxygen.
In the extraction of gold from its ores it was known that dilute solutions of potassium cyanide acted more satisfactorily than concentrated ones. Moreover, L. Janin had noticed that silver is more soluble in dilute than in concentrated solutions. No satisfactory explanation was, however, given.
It seemed, therefore, that an answer was required to the following questions:—
Is oxygen necessary for the solution of gold in potassium-cyanide solutions?
[Footnote] * Mines Report, 1895, pr. 186–189.
If so, does Elsner's equation represent the true amount of oxygen required ?
What is the relation between the concentration of the solution and the rate of solution of the gold; and what is the explanation of the more rapid action of dilute solutions ?
In seeking for an answer to the first of these questions I made several experiments, using both gold foil and gold paper (filter-paper on which metallic gold is precipitated). A gold plate exposed to the action of a 4-per-cent. solution of potassium cyanide, from which air had been removed as far as possible, lost in twenty-four hours only 0.0002 gram., whilst the same gold plate when exposed to the action of this solution with free access of air lost 0.00835 gram. in the same time. A piece of gold paper containing 0.00002 gram, of gold, and having a distinct pink tint, in the absence of oxygen, did not lose its colour for eight days, whilst, when a similar piece was exposed to the action of the same cyanide solution, saturated with air, the colour faded completely in two minutes. These results can leave no doubt as to the absolute necessity of oxygen in order to bring about the solution of gold.
2. Amount of oxygen required for the solution of gold.—I exposed a weighed gold plate to the action of potassium-cyanide solution, enclosed with a measured volume of oxygen in a suitable vessel, and after standing for two or three days I reweighed the plate and measured the volume of the oxygen remaining. I could thus calculate the weight of oxygen required to dissolve a given weight of gold, and I found, as the mean of four experiments, that this weight was within 5 per cent. of that calculated from Elsner's equation. It is therefore evident that Elsner's equation is correct; or, in other words, two atoms of gold require for solution in potassium cyanide one atom of oxygen.
3. What is the relation between the concentration of the cyanide solution and the rate of solution of the gold, and what is the explanation of the more rapid action of dilute solutions?—After numerous experiments, the following plan was adopted to determine this point : Four circular plates of gold, each 22mm. in diameter, were suspended by cotton so as to hang an inch or two from the bottom of a tall glass vessel holding 500cc. of cyanide. By means of a crank attached to a water-wheel, the plates were raised and lowered about an inch twenty times a minute. The solution was kept saturated with air by appropriate means. The plates were exposed to this action for an hour, and were weighed before and after the experiment, the loss in weight representing the gold dissolved. The results are given in the following table :—
|KCN Grams per 100cc.||Gold dissolved.||KCN Grams per 100cc.||Gold dissolved.||KCN Grams per 100cc.||Gold dissolved.||KCN Grams per 100cc.||Gold dissolved.|
These results show that the rate of solution of gold, in potassium-cyanide solutions gradually increases as the concentration of the solution decreases, reaches a maximum at 0.25-per-cent. solution, and then decreases continuously. Precisely similar results were obtained with silver. In seeking for an explanation of this remarkable variation in the solubility of gold in potassium-cyanide solutions I was led to investigate the solubility of oxygen in such solutions, and obtained the following results :—
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
|Percentage of KCN.||Coefficients of Absorption of Oxygen at 18°C.|
These results show that the solubility of oxygen in cyanide solutions decreases rapidly as the concentration increases, a 50-per-cent. solution dissolving little more than one-tenth of the amount of oxygen absorbed by pure water. It will be seen from the table, already given that the amount of gold dissolved by a 50-per-cent. solution is about one-thirteenth of that dissolved by a solution containing 0.25 per cent. It is therefore evident that the rate of solution of gold in these two solutions is almost proportional to the amount of oxygen contained in each. In Table IV. the relations of the gold to the oxygen are shown under the heading Au/O (found). On considering these it is evident that the solubility of the gold is dependent upon that of oxygen, but that something else interferes with the action. For, if the amount of gold dissolved depended solely on the amount of oxygen in solution, the values for Au/O should be constant; but in the results found
it will be seen that these values differ considerably, gradually decreasing as the concentration increases. Therefore, in more concentrated solutions there is less metal dissolved than the amount of oxygen in solution appears to demand. This points to some retarding action on the motion of the molecules. It seemed probable that viscosity has such a retarding action on the motion of the oxygen molecules in solution, reducing their velocity, and consequently diminishing the number of impacts on the surfaces of the plates in a given time, and so decreasing the amount of gold dissolved. In order to test the validity of this conclusion, the rates of soultion of gold were determined in cyanide solutions rendered more viscous by the addition of various substances, such as sugar and glycerol, which might be assumed to exert no chemical influence on the solubility of the metal. The results are shown in Table III. :—
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
|—||Grams per 100cc.||KCN Grams per 100cc.||Gold dissolved in One Hour.||Oxygen Coefficients of Absorption.||Au/O.|
These results prove very conclusively that the assumption in regard to the retarding action of viscosity was correct.
Suppose now that we consider the number of times in a second a give oxygen molecule strikes a surface. We may assume from the results just given that this will depend on the viscosity coefficient z, or, in other words, will be a function of z. So that, if N be the number, we can write N = a + bz + cz2 + &c., where a, b, and c are independent of z (Maclauruin's theorem); or, since Au/O is dependent upon the number of impacts in unit of time, we can write Au/O = a + bz + cz2 + &c. In order to ascertain if these relations hold good for the values of Au and O found, I determined the coefficients of viscosity of a number of cyanide solutions. The observations were made by Gartenmeister's method (Zeit. Physik. Chem., vi., 524), using Finkener's formula—
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
z = r4πp/8lv − vs/8πgl
In this expression r is the radius and l the length of a capillary tube through which a volume v of the liquid of sp. gr. s flows, under a pressure p in unit of time. In the following table the results found by this method are given under z. Under Au/O are shown the values found, and also those calculated by the aid of Maclaurin's theorem—that is to say, by the formula Au/O = a + bz + &c., where a = 0.33 and b = 0.9. The close agreement of the values found by these two methods is sufficient to prove that the true explanation of the smaller solubility of the gold relatively to the oxygen in the more concentrated solutions is to be found in the greater viscosity of these solutions.
|KCN Grams per 100cc.||Gold dissolved in One Hour.||Oxygen Coefficients of Absorption||.Au/O.|
Similar experiments are recorded in the papers above referred to with regard to silver, and precisely the same results were obtained with this metal.
The following is a summary of results :—
Oxygen is necessary for the solution of gold in potassium cyanide, and no gold is dissolved in its absence.
The ratio of the gold dissolved to the oxygen required for its solution is 196 : 8, as demanded by the equation—
4Au + 8KCN + O2 + 2OH2 = 4AuCNKCN + 4KOH.
The rate of solution of gold in potassium-cyanide solutions varies with the strength of the solution, being small for concentrated solutions, increasing as the solution becomes more dilute, reaching a maximum at 0.25 per cent, of cyanide, and then again diminishing.
The rate of solution of silver in potassium cyanide varies in the same way, and the maximum is reached at the same degree of dilution.
The ratio of the amount of gold dissolved by any given cyanide solution to that of the silver dissolved by the same solution is nearly the ratio of their atomic weights.
The variation in the rate of solution of gold in cyanide solutions is not directly influenced by the amount of cyanide in solution, except in the case of very dilute solutions, but is mainly due to the solubility of oxygen in these solutions, the amount of gold dissolved being nearly proportional to the absorption coefficients of oxygen in such solutions.
The rate of solution of gold is, however, not exactly proportional to the above-mentioned coefficients, but is rather less than it should be for the more concentrated solutions.
The explanation of this diminishing ratio of the gold dissolved to the oxygen available, as the concentration of the solution increases, is to be found in the increasing viscosity of the solutions as the quantity of cyanide augments.
The explanations given in 6, 7, and 8 are equally applicable to the solution of silver in potassium-cyanide solutions.
Returning to Mr. Skey's paper, to which I have already referred, I give the following extract in order to show the position he takes up. He says, “Why very weak cyanide solutions act as swiftly as they do, while strong solutions do not act upon gold to a degree or at a speed in any way corresponding to what we expect, is a problem that has not, I think, been solved. To account for this it has been assumed that strong solutions of the cyanide do not dissolve oxygen or are not permeated by it as readily as weak solutions are. But that there is a plentiful supply of oxygen in these solutions is made manifest by the results of the following experiments :—
“1. A newly-made cyanide solution of greatest strength is poured into a shallow vessel, and at the bottom of it a small slip of gold leaf gummed on paper is placed. A long slip of the same is then placed so that one end rests upon the bottom of the vessel while the other end projects out of the solution. In a few minutes it may be seen that the whole of the gold on the long slip has been dissolved, while the piece that is wholly immersed in the fluid does not appear to be at all affected.
“2. In the same solution place a slip of gold leaf coupled with platina, so as to lie also at the bottom of the vessel, when in a shorty time it may be shown that the gold has entirely dissolved, while the gold leaf that was not paired with any negative substance has not been affected. With chalcopyrites for the negative pole the solution of the gold was far more rapid than in the former experiment, when platinum was used for this purpose, showing the advantage there is in pairing the gold with a substance that is strongly electro-negative to it.
“I think the results of these experiments clearly prove
that there is a sufficiency of oxygen present, even in the strongest solutions of potassic cyanide, to allow of the rapid solution of gold therein. But why strong cyanide solutions have so little, or so very slow, an effect upon gold as we find is a question that, in the light of these results, appears as yet quite unanswered. For my part, I am inclined to think that a compound forms upon the gold when in strong cyanide solutions that is either insoluble or but very slowly soluble in these strong solutions, but is soluble to a considerable extent in weak solutions. It is, I think, very probable that the cyanide of gold that first forms on the gold has to be dissolved as a simple cyanide before it can be so acted upon by the potassic cyanide as to pass into the comparatively-soluble aurocyanide of potassium.”
It is remarkable that Mr. Skey should make the statement, “It has been assumed that strong solutions of cyanide do not dissolve oxygen, or are not permeated by it as readily as weak solutions are,” when he had before him my papers proving that oxygen is less soluble in concentrated than in dilute solutions. In the two experiments that follow this statement Mr. Skey shows that gold when partly immersed in the solution or when coupled with platinum is rapidly dissolved, and concludes “that there is a sufficiency of oxygen present, even in the strongest solutions of potassic cyanide, to allow of the rapid solution of gold therein.” It may be as well in the first place to point out that these experiments do not deal with the question of concentration in a satisfactory manner. Whilst it is shown that by partial immersion or contact with platinum the rate of solution of the gold is greatly increased, nothing is given to show the relative effect of solutions of varying concentration in such cases. It is unnecessary, however, to consider this point further at present, as I shall show that in these particular cases the solution of the gold is not due, as Mr. Skey assumes, to the oxygen in solution, but to electrolytic action.
Perhaps it will be best to consider the question in the successive steps which I took in my investigation.
Strips of gold leaf (1 to 3, Fig. 1), gummed on paper, were gummed on to the inside of a beaker, into which a saturated solution of potassium cyanide was then poured, until the top of No. 3 strip was just covered. In five minutes No. 2 was almost entirely dissolved, whilst No. 1, which extended partly over the bottom of the beaker, required about thirty minutes. The top of No. 3 was only about 1/6in. beneath the surface of the liquid, and yet after standing for an hour there was no alteration in its appearance. Nos. 2
and 2a were separated by a space of about 1/10in., but No. 2a showed no change in an hour.
These results prove that for rapid solution a portion of the strip must project above the surface of the liquid; they also suggest the probability of the solution being due to electrical currents maintained by the action, at the surface of the liquid, of cyanide and the oxygen of the air.
In pursuance of this idea the following experiments were made: The strips shown in Fig. 2 (varnished as shown by the shaded portions were exposed to the solution used in the last experiments. In half an hour that portion of No. 6 below the surface was completely dissolved. In No. 5 the action was the same—all the strip beneath the surface of the liquid, with the exception of the varnished portion, being dissolved. The same was the case with 7, both a and b dissolving, and apparently at about the same rate. No. 4 showed no alteration. It was suggested above that the solution of the lower portions of the strips was due to electrical action, and this is clearly proved to be the case by the results obtained with Nos. 5 and 7 strips. Let us consider how No. 5 differs from Nos. 2 and 2a in the last experiments. So far as chemical action is concerned, one would naturally conclude that 2a should be dissolved rather than 5a, since the varnished portion separating 5 and 5a is very much wider than the space between 2 and 2a, and for such action a portion protected by varnish is quite as great a barrier as a blank space. When we consider electrical action, however, the case is quite different : the space between 2 and 2a does not allow an electric current to pass, whilst the varnished portion between 5 and 5a offers no obstruction to such a current. It is therefore most probable that a current flows, and that the solution of 5a is due to its action. Again, in No. 7 the solution of b can be explained only by the generation of electric currents in a. Hence it follows that the solution of the lower portion of any strip partly immersed in concentrated cyanide of potassium is due to electrical action.
Should any doubt as to the production of an electric current remain, the results obtained in the following experiment must remove it :—
Fig. 3 shows two plates of gold, varnished as shown by the shading. The upper ends of these plates were connected through a Thomson's galvanometer (G in the figure). A rested
on the bottom of a beaker partly filled with a saturated solution of potassium-cyanide, the surface of which stood about half-way up the varnished portion of this plate. As previously shown, with strips of gold paper varnished in a similar manner there will be practically no action on A. When, however, No. 2 plate is lowered into the solution so that B is partly immersed there is strong deflection of the galvanometer, proving that a considerable current is passing; and, from the direction of the deflection, showing that cyanogen is being deposited electrolytically upon A, and that therefore aurous cyanide is formed, which dissolves in the liquid. Hence the gold plate A is dissolved. When No. 2 plate is further lowered, so as to have the surface of the liquid at about the middle of the varnished strip C, the deflection is hardly perceptible; whilst on still further lowering the plate, so as to expose part of D to the action of the solution, the deflection is the same as at first.
Let us consider the course of the current in the case of a straight strip of gold partly immersed. From the experiments just given it is evident that there must be an electromotive force at the surface of the liquid, and that a current will flow down through the metal and back through the solution to its starting-point. We may consider the portion of the strip beneath the surface of the solution as one pole of a battery, and that portion just at the surface as the other pole. Now, it is well known that when a current passes through an electrolyte (in this case potassium-cyanide solution) the latter is decomposed, its constituents—the ions—appearing at the poles. In the present case cyanogen is liberated at the lower pole, and at once combines with the gold to form aurous cyanide, which unites with potassium cyanide in the solution to form the soluble double salt; whilst the potassium must pass to the upper pole, where it will either decompose water, liberating hydrogen, or unite at the moment of liberation (whilst nascent) with the oxygen at the surface. In either case potash will be formed at the surface.
In order to prove that this is the case, the following experiments were made:—
Three test-tubes (shown in Fig. 4) of the same size were taken. Into No. 1 test-tube 10cc. of a 2-per-cent. solution of potassium cyanide almost free from air were poured. A gold plate was dropped in, and then a thin disc of cork, through the centre of which passed a rod of gold, was placed as shown in the figure. The disc of cork fitted the tube tightly, but a very small strip was cut away at
one side so as to allow all the air to be driven out when the cork was forced into its place. A second 10cc. of the same solution were now added. The gold rod passing through the cork was varnished from a point a little below the surface of the liquid to within ⅛in. of its lower end, which rested on the gold plate. No. 2 test-tube was filled up in exactly the same way, whilst No. 3 differed only in having neither gold rod nor plate. The three tubes were placed in a small beaker standing in water, and were then covered by inverting over them a beaker, which also dipped into the water, and so protected the solutions from any fumes that might be in the laboratory. After standing for five days the solutions in the upper and lower portions of the tubes were removed separately and analysed in the following manner: 5cc.—i.e., half of the solution—from the top or bottom of each test-tube was titrated with AgNO3N/50. This gave the amount of free potassium cyanide remaining in the solution; the other half was evaporated to dryness with the addition of a few drops of sulphuric acid, and the residue was heated first alone and then with small additions of ammonium carbonate. This residue, which consisted of potassium sulphate and metallic gold, was weighed, and the gold determined by the usual assay process. The amount of potassium sulphate obtained from 5cc. of solution, both above and below the cork, was thus known. From the gold found the amount of potassium cyanide in the double cyanide of gold and potassium can be calculated; the amount of free potassium cyanide is also known. The sum of these two values, calculated as sulphates and subtracted from the total amount of potassium sulphate found, gives the amount of potassium sulphate existing in the solution as potash. Now, if the solution of the plate be due to oxygen in the solution below the cork, potash will be formed there; therefore, in such a case, the ratio of the potash found below the cork to that found above it will be the ratio of the gold dissolved in these two places. If, however, the solution of the plate be due to an electric current generated at the surface of the solution, no potash will be formed below the cork, but only at the surface of the solution. Of course, there is sure to be some oxygen below the cork, and this will act upon the plate independently of the electrical action, and so form a small amount of potash in the lower portion of the tube. However, to prove that a current flows, and in the direction stated above, it is only
necessary to show that the ratio of the potash formed above the cork to that formed below it is greater than the ratio of the gold dissolved above to that dissolved below. The following results show that this is the case :—
|(1.) Gram.||(2.) Gram.|
|KHO from 5cc., top of test-tube||0.0155||0.0171|
|" " bottom "||0.0075||0.0088|
|Gold " top "||0.0333||0.03195|
|" " bottom "||0.02674||0.00232|
|KHO above cork/KHO below cork =||0.0155/0.0075||0.0171/0.0088|
|Gold above cork/Gold below cork =||0.0333/0.02674||0.03195/0.0232|
The titration of the solutions in No. 3 test-tube showed that, although the decomposition in the top solution was rather greater than that in the solution below the cork, the difference was so slight as not to affect the above results.
These experiments clearly show that the solution of a gold plate partially exposed to the air is largely due to electric currents acting in the manner above described.
In regard to Mr. Skey's second experiment, in which he coupled gold lying at the bottom of a cyanide solution with platinum, I have repeated it several times, but have failed to get his results. This may be due to the fact that he does not enter into detail sufficiently to insure any one repeating the experiment under exactly the same conditions. The following are some of my experiments in regard to this matter:—
In Fig. 5, 1 and 2 are small strips of gold foil (gummed on paper) lying at the bottom of a saturated solution of potassium cyanide. 3 and 4 are like the last, and in addition they have lying across them narrow strips of platinum attached with gum to one edge of the gold. 5 and 6 differ from 3 and
4 only in having a platinum, wire in 5, and a strip of platinum in 6, resting as shown in the figure, and projecting above the surface of the liquid. In half an hour 1, 2, 3, and 4 were apparently unaltered, whilst 5 and 6 were almost completely dissolved. After twenty-four hours 1, 2, 3, and 4 showed no appreciable alteration. I now suspended a platinum wire so that one end rested on No. 1 strip, the other projecting above the surface of the liquid. In the same way a platinum strip (varnished at the surface of the liquid) was suspended so as to touch No. 2 strip. In forty minutes No. 1 was almost entirely dissolved, whilst No. 2 showed no alteration.
These results are very similar to those obtained with gold strips partly immersed, and show clearly that, as in the former case, solution of the gold is due to an electric current generated at the surface of the liquid.
The following experiments were made in order to determine what influence oxygen in solution has on the electrical solution of gold. Two circular plates of gold of the same diameter were marked 1 and 2 respectively; a narrow slit was cut in each near the margin, and the plates weighed. Through one of these slits one end of a long strip of gold about ¼in. wide was passed and bent back on itself, being tapped with a small hammer to insure close contact with the plate; the other end of the strip of gold was bent round a glass rod, resting across the top of a Nessler glass containing 50cc. of cyanide solution. By this means the plate was suspended in the cyanide solution at about ½in. from the bottom of the vessel. The gold strip was varnished from within ½in. of the plate to within ½in. of the surface of the liquid. The second plate was suspended by cotton at the same depth in the second Nessler glass, which also contained 50cc. of the same cyanide solution. In other words, the conditions were the same for the two plates, except that one was suspended by gold and the other by cotton. After being exposed to the action of the cyanide solutions for one hour the plates were removed and reweighed. The strip of gold was now attached to No. 2 plate, No. 1 being suspended by cotton, and the experiment repeated, still keeping No. 2 plate in No. 2 Nessler glass. These four determinations were made first in a solution containing little air, and then in the same solution saturated with air. The results are shown in the following table, the means only being given. The weights shown are in terms of assay weights : 1,000 = 0.5 gram :—
|KCN per Cent.||Gold Plate suspended by||With Little Air in Solution.||Solution saturated with Air.||Increase due to Air.|
If we consider the totals in the above table it will be seen that in the case of the plates suspended by cotton saturation of the solution with air increases the amount of gold dissolved by 3.35, whilst in the case of the plates suspended by gold the increase is only 2.95. In the former case, solution takes place in the manner already explained in my earlier papers on this subject; in the latter, in addition to this mode of solution there is solution due to electrical action. Now, if these actions be independent of each other, the results found for the plates suspended by gold must represent the sum of these actions (chemical and electrical). Suppose this to be the case: then, in the above table, the amounts (totals) dissolved by electrical action will be 2.65 and 2.25—i.e., 5.65 - 3 and 8.6 - 6.35. It appears from this that an increased amount of oxygen in solution decreases the electrical action, whilst it very materially increases the chemical action. Of course, the determinations made are not numerous enough to insure absolutely correct results, but they are quite sufficient to show that the electrical action is independent of the oxygen in solution. Therefore, Mr. Skey's conclusion, that because gold partially immersed in saturated cyanide solutions rapidly dissolved there must be a considerable amount of dissolved oxygen, is incorrect.
After stating this conclusion, Mr. Skey proceeds—“But why strong cyanide solutions have so little or so very slow an effect upon gold as we find is a question that in the light of these results appears as yet quite unanswered. For my part, I am inclined to think that a compound forms upon the gold when in strong cyanide solutions that is either insoluble or very slowly soluble in these strong solutions, but is soluble to a considerable extent in weak solutions. It is, I think, very probable that the cyanide of gold that first forms on the gold has to be dissolved as a simple cyanide before it can be so acted upon by the potassic cyanide as to pass into the comparatively-soluble aurocyanide of potassium.”
As my experiments have fully answered the points raised by Mr. Skey, it is hardly necessary to consider the above hypothesis except to point out that there is nothing to base it upon. I have already shown that in potassium-cyanide solutions the rate of solution of gold is due to the same causes as that of silver. Now, it is well known that silver cyanide is more rapidly soluble in concentrated than in dilute solutions of potassium cyanide, and why Mr. Skey should assume that the reverse is the case with the gold compound is inexplicable.
In conclusion, I wish to point out that nothing has been adduced by Mr. Skey to weaken the theories stated by me in regard to the solution of gold in potassium-cyanide solutions. These theories were advanced to explain the action of pure potassium-cyanide solutions on pure gold wholly immersed therein, and of course were never intended to include the cases of partial immersion, or those of contact with other substances, both of which require special consideration of the kind described in the latter part of this paper.