And interesting study of this problem, with a practical method of increasing the oxygen content of cyanide solutions, is presented an article by G. K. Prentice, in the journal of the Chemical, Metallurgical, and Mining Society of South Africa . The following is an abstract;

The usually accepted reaction is 4 Au + 8KCN + 20 + 2H2O = 4KAu(CN)2 + 4KOH,according to which 521 parts of KCN require thirty-two parts of oxygen for the satisfactory dissolution of gold. Thus a solution with 0.05 percent KCN should contain 30.7 mg of oxygen per litre and for maximum dissolving efficiency, and a 0.02% solution, such as used as at Modderfontein East, should contain 12.3 mg per litre. One reason for the apparent lack of interest in this problem is, that prior to 1918, when H. A. White introduced his method for determining the oxygen in cyanide solution, based on the colour imparted to alkaline pyrogallol, a detailed knowledge of the oxygen content of all solutions in circuits had been difficult to acquire.

At the altitude of the Rand, with a normal barometric pressure of 620 mm, about 5 mg of oxygen should dissolve in a litre of water and 15 degrees Celsius on exposure to the atmosphere.

Assuming that cyanide solutions will dissolve about the same quantity, one may expect at least six mg of oxygen in the solutions on the Witwatersrand . Although in exceptional cases some solutions do come up to this figure, many show a considerable deficiency in oxygen, and one of the problems associated with the dissolving of gold is to account for this deficiency and to find a remedy therefor. In the Central Mining Rand Mines group the oxygen content varies from 2.5 to 5.5 mg., with the average below 4 mg.

Here the questions naturally arrives: Why do not these solutions, so freely exposed to the atmosphere, carry anymore oxygen, and by what means can the oxygen content the increased? This problem became acute at the Norse mine in 1931, when the oxygen content dropped to one mg, with a corresponding long sequence of high residues. Undoubtedly the principal cause for this deficiency may be traced to reducing matter in the ore and in the mine water. Here it should be noted that the deficiency is the greatest in those mines where neutralised mine water is the chief supply for mill water, particularly if the settlement area after neutralisation is limited. Several methods for raising oxygen content may here me mentioned, such as the generation of nascent oxygen in the solution by electrolysis, the addition of oxidising agents, or the use of atmospheric oxygen.

According to Henry's Law, the weight of a gas absorbed by a given volume of liquid is directly proportional to the pressure of the gas. If at one atmosphere six mg of oxygen is absorbed, at 5 atmospheres this figure would be 30 mg., which, as previously stated, is the theoretical quantity required to give maximum efficiency with a solution containing 0.05 percent KCN. Obviously, when the pressure is released, the oxygen content again diminishes, but no matter how great the deficiency before the pressure is applied, it may be assumed that the excess oxygen, while under pressure, destroys the reducing matter present; so that on reversion to atmospheric pressure the oxygen will not again fall below six mg.

To test the correctness of these deductions a large quantity of solution at the Norse mind, containing 2.5 mg of oxygen, was subjected to various kinds of treatment to raise the oxygen content. A. review of these laboratory tests indicated that little improvement was obtained by bubbling compressed air through the solution at atmospheric pressure. Introduction of the air by a mechanical mixer increased the dissolved oxygen to a maximum of five mg., with a consequent acceleration in dissolving, but such intimate contact between air and solution is hardly economical on a large-scale, on account of high consumption of power and cyanide. Addition of pure oxygen gas is also too costly. Keeping the solution in contact with compressed air for 20 minutes gave a solution saturated with oxygen of high dissolving efficiency.

As the latter procedure was considered sufficiently attractive to justify further investigation, a small semi-commercial unit was put into operation, in which the solution was passed through a cylinder of three foot diameter and 5.17 ft long, the capacity of which, when half fall, was 0.42 tonnes, with an exposed solution area of 15.5 square feet. Solution was fed through a pipe with 20 1/2 inch holes, placed in a row to throw the solution upward. Compressed air was added to the top through a 1/2 inch pipe, and a 2 in discharge pipe was placed at the bottom. Several tests were made with an air pressure averaging 65 pound. Initial oxygen conte are nt was 3.5 mg., that of the discharge 6.5 mg., when a rate of flow of 0.42 tonnes per minute, solution remaining in the cylinder one-minute.

Upon discharge the solution was milky white from dispersed air, but it cleared after one minute's standing; oxygen tests were made on the clear solution. Air consumption was negligible, less than 2 cubic feet of free air per minute, and cyanide consumption was likewise negligible. A slight escape of the air from the cylinder was found desirable to prevent concentration of nitrogen. Solution having six mg of oxygen, when left in an open bucket, showed five mg after one hour, 4.5 mg after five hours, and 4 mg after 22 hours. Left in the cylinder 16 hours with 100 pound pressure the solution showed seven mg upon release.

These results prove conclusively that the oxygen content of mill solutions may be materially improved by this simple, continuous process. The next step in the investigation was to ascertain what benefit would be derived by treating ore on a large-scale with such a solution. An acceleration in the dissolution of gold was expected, especially the absence of periods of erratic results such as experienced in nearly all cyanide plants, although no appreciable improvement over normally good residues was looked for. Some reduction in cyanide consumption was also anticipated. Six 410 tonne sand charges were treated, three in the normal manner and three with aerated solution; with the equipment available it was not possible to use more than an average of 36 percent of aerated solution on the latter. Final residues in aerated treatment assayed 0.34, 0.33, and 0.335 dwt.gold per ton, compared with 0.35, 0.375, and 0.345 dwt. for the normal charges. The most remarkable result was perhaps the pronounced reduction in cyanide consumption, an average of 49.4 pound sodium cyanide for the aerated solution against 75 pound for the normal treatment.