Attempts to discover, for the cyanide process, a better precipitant than zinc have been unsuccessful. Hydrogen sulphide, sulphurous acid, ferrous sulphate, etc., precipitate gold and silver from cyanide solutions either not all or incompletely.

Aluminum has been tried, but it costs too much, and cannot be used in solutions containing lime. Charcoal has been used to a small extent, in treating acid tailings, the solutions of which were acid or very weakly alkaline. It involves an excessive cyanide-consumption; must be used in large bulk; and subsequently all burned to ash, with risk of losing gold and silver. On the other hand, the ash is easily melted, almost fluxing itself; and the bullion is fine.

Of the many proposed electrical methods of precipitation, only the Siemens-Halske or some modification of it has proved successful. But this method is losing rather than gaining ground. Mr. John Yates, in a paper comparing the zinc and the electrical methods of precipitation, shows that the former presents decided advantages. In fact, zinc may be said to take the lead more and more, as its use is better understood.

One of the weak points in this method has been the refining of the zinc-precipitates (containing from 4 to 50 per cent, of bullion, and from 20 to 40 per cent, of zinc, together with base metals, metalloids, silica, lime, etc.), from which it is difficult to obtain clean bullion without loss.

Zinc is used as a precipitant in two forms:—shavings or filiform zinc, and zinc-dust. The latter, first proposed, I think in connection with the bromo-cyanide process of Sulman and Teed, did not win much favor until it was adopted by C. W. Merrill, at Marysville, Montana, and later by the Golden Gate mill at Mercur, Utah. Lately, a number of large American works have adopted it. Its general use is hindered by the difficulty of getting zinc-dust of uniform quality and free from oxide, and by the circumstance that, for some unexplained reason, zinc-dust precipitates do not yield as readily to acid-treatment as do those from zinc-shavings.

Nature of Precipitates

Whether from zinc-shavings or dust, the precipitates are apparently alike in general composition. Mr. William Magenau gives analyses of both, which support this statement.

As to the forms or combinations of the various constituents, it has been generally assumed by the authorities that the metals, gold, silver, copper, etc., are present in the metallic state. A few writers have suggested that the black precipitate may be an intimate combination of gold with something else. Gage thinks it is a double cyanide of gold and zinc. Rose says:

“ Pure zinc has only a slower action on solutions of potassium cyanide than ordinary commercial zinc containing about one per cent, of lead ; and the action of the gold-zinc couple, formed by the black deposit of gold (which may be really a compound of gold and zinc) and the unaltered zinc is still more vigorous.”

The following are the conditions of practice under which gold and silver are precipitated by zinc from cyanide solutions:

In a slightly alkaline cyanide solution, containing usually less than 1 oz. of gold and silver per ton (2,000 lb.) of solution, and in many cases as low as from 0.05 to 0.02 oz. (and even lower in the case of final washes), or 1 part of gold and silver in from 24,000 to 1,200,000 parts of solution, the gold and silver are precipitated by the strongly positive metal, zinc, which is in a finely divided state and hence very active chemically.

F. Mylius and O. Fromm have made an interesting series of experiments upon very dilute solutions of the salts of different metals,—silver, gold, copper, etc. Upon immersing a zinc rod in the solutions of these metals they found that the metal in solution was precipitated as a spongy black mass, which they determined to be a definite alloy of zinc and the metal. As a result of their investigation they concluded (among other things) that the metals are capable in the moment of their separation, and at ordinary temperatures, of combining with each other; and that, by the action of positive metals upon dilute solutions of negative metals, alloys are electrolytically formed.

In the case of the silver-zinc alloy, it was found that if the strength of solution went above 1 part of silver in about 33,000 parts of solution, white crystalline metallic silver was precipitated. If less silver was in solution than 1 part in 33,000, the black precipitate or alloy of silver and zinc was formed.

The degree of dilution necessary to form these alloys probably differs with the different metals; and it also seems possible that there is a whole series of these alloys formed by the action of zinc in solutions of varying metal-content. All the experiments of Mylius and Fromm were made with neutral or slightly acid solutions.

Are such alloys formed when gold and silver are precipitated from a cyanide solution by zinc ? In that case, the solution is slightly alkaline, and the proportion of metal is about from 1: 24,000 to 1 : 1,200,000.

It is well known that gold and silver precipitated from a cyanide solution (not containing too much of these metals) come down as the black flocculent precipitate, looking like that obtained by Mylius and Fromm.

This precipitate, after very fine sifting, which undoubtedly removes most of the free zinc, still contains a high percentage of zinc, although but little can be seen by the microscope. Moreover, when the metal-content of the solution is above a certain point, the zinc assumes the color of gold or silver, as the case may be; and the gold or silver is precipitated in the metallic state, or a near approach to it.

Philip Argall says:

“When the solutions pass 1.5 oz. per ton the gold on the shavings assumes a yellow color, and passing 2 oz. and upward is usually golden yellow in the first compartment, shading off in the following compartments to the usual black- colored deposit that collects on the zinc in ordinary precipitation from poor solutions.’’

This passage describes in fact the appearance of the gold precipitated by zinc from solutions of varying gold-content, the other conditions remaining almost the same. The solution upon entering the precipitation-box contains its maximum gold-content (in the above case, 2 oz., and upwards), but as it progresses through the several compartments of the precipitation-box, and the gold is precipitated, the solution becomes poorer and poorer in gold, and the character of the precipitate is changed. In the first compartment the gold is probably in the metallic state or a near approach to it. In the second is a compound containing more zinc, and so on throughout the series until we finally reach an alloy, containing the maximum percentage of zinc.

Experimental Work Upon Alloys

In order to investigate the formation of zinc-gold, and zinc-silver alloys from cyanide solutions, a simple apparatus was fitted up, consisting of two ordinary oil-barrels, as tanks, placed with the top of the second lower than the bottom of the first, so that a current of solution could be passed from one to the other through a series of four interposed glass bottles, containing the zinc precipitant. The shavings having been placed in the bottles, the silver- or gold-solutions were caused to pass very slowly through them.

Silver

The first experiments were made with 400-lb. samples of the following solutions:

In each case, the silver was deposited in a hard, thin, adherent coating, which could not be removed for separate analysis. It varied in color from light straw-yellow to copper-red, except that, in case I., it was from gray to black in spots where the shavings were somewhat rough. In cases I. and II., a considerable quantity of hydrogen (recognized by collecting and burning it) was evolved. In III., there was very little hydrogen, and the action generally was much slower. As already observed, the precipitate could not be removed for examination or analysis. But I have found an alloy containing from 40 to 50 per cent, of zinc and the balance of silver to show much the same characteristics. The rough surface of this alloy has the same copper-red color, and even the polished surface has a very distinct reddish tinge. With acetic acid, little or no action takes place in either case. Other reagents, such as chromic acid, hydrogen peroxide, etc., give the same reactions in both cases.

That this coating is more energetic than zinc as a precipitant of gold, copper and silver, may be seen by watching the precipitating-bottles. At first the precipitation is very slow; but after the shavings have assumed the copper-red hue it goes on much more rapidly.

Gold

A similar experiment was tried with a cyanide solution containing 0.2 per cent, or 4 lb. KCN per ton, 025 oz. gold and 1 lb. NaOH per ton. When this solution was passed through them, the zinc shavings rapidly assumed a dark purple- black color, which deepened as the process went on, until, as the precipitate became heavier, it detached itself from the zinc and fell to the bottom of the bottle.

To avoid oxidation, this fine precipitate was dried in an atmosphere of illuminating gas, after which it was shaken lightly on a 100-mesh screen to remove particles of metallic zinc. The resulting product was found to contain 40.11 per cent, of zinc, the rest being probably gold. It was found to oxidize very rapidly in the air. I have noticed the same phenomenon in commercial practice. When zinc shavings pretty well charged with precipitate are removed from the zinc-box and exposed to the air, they become quite warm through rapid oxidation.

For purposes of comparison, an alloy of gold and zinc, containing from 40 to 50 per cent, of zinc and the rest gold, was prepared in the ordinary way. It formed a hard white mass, rapidly acted upon by nitric acid; metallic gold being set free. In a finely divided state, it is also a very energetic precipitant of gold.

The purple-black precipitate, when subjected to pressure under an agate pestle, assumed the same appearance; and it acted in the same way with nitric acid and other reagents.

A portion of the precipitate allowed to remain in the precipitation-bottles after all of the free zinc had been consumed, still continued to precipitate gold with evolution of hydrogen. At first the action was very energetic; but as the zinc in the precipitate decreased, it became less and less active until it apparently ceased altogether.

The purple precipitate first shaded off into a dark brown, then through various shades of yellow until it assumed a metallic luster and a golden color. In this last stage it does not seem to contain much zinc; the exact amount, I have been unable to determine.

Conclusions.—It may be inferred from these experiments:

1. That the precipitates formed by zinc in the cyanide process are probably alloys or intimate mixtures of zinc and the metals in solution; and
2. That they follow certain regular laws as to their formation and composition.

Methods of Refining

The various metals and elements may be considered as occurring in the following combinations: gold, silver and copper, as intimate mixtures or alloys with zinc; mercury, as an amalgam, probably of zinc; lead, arsenic and antimony, in unknown combinations; iron and aluminum, as hydroxide, oxide or silicate, due to clay; calcium, as sulphate, carbonate or hydroxide; magnesium, probably as carbonate; silicon, as silica (sand) and silicates; zinc:

• (a) in metallic form, due to excess of zinc used in precipitation,
• (b) as oxide, due to oxidation during handling of the precipitates, and
• (c) in combination with gold, silver, copper, etc., as intimate mixtures or alloys.

Cadmium, resembling zinc in its chemical behavior and being a strongly positive metal, probably forms similar alloys when present in the zinc used for precipitation.

Finally, there are in the precipitates complex organic com-pounds, of doubtful nature. The numerous impurities mentioned above, coming from the ore, the zinc (which is, in practice, never pure), and the lime used for neutralizing acidity of ore, render the refining difficult.

There are three general methods of refining:

1. roasting, with or without an oxidizing agent (such as niter), then smelting oxidized precipitates with suitable fluxes in graphite crucibles;
2. acid-refining, which consists of preliminary treatment with some acid, usually sulphuric acid, then smelting acid- treated precipitates with proper fluxes in graphite crucibles; and
3. smelting with lead, either with or without preliminary acid-treatment.

Refining by Roasting

This method has been more or less used since the introduction of the cyanide process. At one time it was the general practice in South Africa; but it has never prevailed in the United States.

The roasting of the precipitates is usually done in cast-iron muffles, heated with coal or wood. The precipitates are spread out in thin layers upon large iron trays, and stirred from time to time during the roasting. Roasting in pans, heated from the bottom by an open furnace, is not considered as good practice as roasting in a muffle. In many cases, oxidation is aided by the use of niter, applied to the precipitates as a strong solution or mixed with them in the dry state; from 3 to 5 per cent, is the usual amount used; any excess over the quantity necessary to oxidize the base metals present is to be avoided, as it rapidly destroys the graphite crucible used for melting. The oxidized precipitates are mixed with suitable fluxes and melted in graphite crucibles.

The main object is a thorough oxidation of all base metals, so that in the subsequent melting they may be carried into the slag.

The advantages of this method are,—the simplicity of operation and the requirement of no acids or other chemicals aside from fluxes.

Its disadvantages are,—the production of low-grade bullion; the heavy losses of precious metal; the difficulty of roasting properly; and the rapid corrosion by the strongly oxidizing charge of the graphite crucibles used for melting,—especially when niter is employed.

These drawbacks are so serious that the method is rapidly going out of use.

Refining with Acid

Sulphuric acid only has been used to any great extent for the preliminary treatment of the precipitates. Nitric acid has not proved satisfactory. Hydrochloric acid seems to give as good results as sulphuric, but costs more, and might possibly occasion the evolution of chlorine, which would dissolve gold. Acetic acid has been suggested, but I have found no record of its use in practice.

The acid-treatment is usually carried on in wooden tanks, sometimes lined with sheet-lead. At some plants, the mixed precipitates and acid are heated in the tank by steam introduced through the mass near the bottom. At some of the older works this heating is done in a large leaden dish supported over a furnace. In many places, however, the heat generated by the mixture of the concentrated acid and water is deemed sufficient. Mechanical stirrers are used, but in many cases the mixture is stirred by hand with a wooden paddle.

The strength of acid varies greatly. In general it is safe to say that the best results are obtained by using rather dilute acid (1 part of concentrated acid to from 8 to 10 parts of water), at least for the first treatment. If the acid is too strong, there is greater danger of the formation of the less soluble base-metal salts, which cannot be completely removed by washing ; this largely defeats the removal of zinc, which is the chief purpose of the acid-treatment.

In small and poorly-equipped plants, the precipitates are allowed to settle and the clear solution siphoned off, wash-water is then added and the mass again allowed to settle and the clear solution siphoned off. This is repeated until the precipitates have been completely washed. This procedure takes much time; zinc sulphate can never be completely removed in this way; and serious losses may arise from fine particles of precipitate going over with the wash-water.

In the best practice a vacuum-filter or a filter-press is used. By the latter method a rapid and complete separation of the solid matter from the liquid may be made with but very little loss.

After the zinc is in solution, the important point is the complete removal of zinc sulphate by thorough washing. The thoroughly-washed precipitates are best dried in a cast-iron muffle heated from the top only, which prevents loss through boiling and spattering. (In some cases a roast is given at this point.) The material is then mixed with fluxes (borax, soda, sand, fluorspar, etc.,) in such proportions as to give, with all of the impurities and base metals, a fluid slag that does not contain shots of bullion. The melting is done in graphite crucibles, as in refining by roasting. C. Butters has used a reverberatory furnace for this purpose.

The advantages of acid-refining are, that it produces fairly fine bullion (the fineness usually depending upon the care exercised in conducting the operation), and that the losses are not so heavy as in the roasting process. Its disadvantages are, that it calls for a considerable equipment and the losses due to handling are comparatively large.

With careful work the loss is usually said to be about 1 per cent, of the value, but there is little doubt that, in most cases, it is larger. Mr. C. W. Merrill states the losses in ordinary methods of refining (meaning, probably acid-refining) at from 2 to 6 per cent. Alfred James gives the losses by the roasting method as from 0.5 to 6 per cent., and adds:

“Acid treatment yielded results varying with the kind of acid used, nitric acid showing a greater loss than hydrochloric and sulphuric. On the other hand, with impure zinc, that in ordinary use, nitric acid gave the purest bullion, but with lead-free zinc, dilute sulphuric acid gave both the purest bullion and the lowest acid loss.’’

One trouble in acid-refining is the formation of poisonous gases when the acid attacks the precipitate. Any cyanide remaining in the latter will thus form the deadly hydrocyanic acid. Arsenic (which may have been in the ore or in the zinc) forms hydrogen arsenide, which is fatal if inhaled. Several men have lost their lives by breathing hydrogen arsenide at the Golden Gate mill, Mercur, Utah, where ores containing arsenic are treated. The danger was finally overcome at this mill by using nitric acid in connection with the sulphuric acid (about one part of nitric to two of sulphuric), and ventilating thoroughly with exhaust-fans. The nitric acid oxidizes hydro-cyanic acid to harmless cyanic acid, and the hydrogen arsenide to arsenic acid. Moreover, it aids in the solution of zinc, which is much more soluble in sulphuric acid in the presence of an oxidizing agent.

In some plants the precipitates are roasted before being treated with acid. This generally results in the more complete removal of zinc and other impurities, but the expense and probable loss involved in this preliminary step make its general employment scarcely advisable. As already remarked, precipitates formed by zinc dust are less completely acted upon by acid than those formed by zinc shavings. An instance in point came to my notice in the Black Hills, at a mill using zinc dust. When dilute acid was added to the precipitates in the acid-tank, little if any action could be observed. The mixture was left in the tank for a week, being agitated and warmed from time to time by blowing steam through the mass. At the end of the week, it was filtered, and found to contain almost as much zinc as at the start. This experience led to the substitution of zinc shavings at the mill in question.

At the Golden Gate mill at Mercur, Utah, on the other hand, zinc dust is used as a precipitant, and no such trouble is encountered with the acid-treatment. The preliminary roast, and the use of nitric acid, already mentioned, probably solve the problem in this case. It would be interesting to know what percentage of zinc remains in the acid-treated material.

At the Homestake Mining Co., Lead, S. D., zinc dust is likewise used as a precipitant; but although the acid-treatment is thorough, it leaves in the precipitates from 4 to 12 per cent, of zinc.

Several causes may contribute to this partial failure of the acid to dissolve the zinc. As already observed, the remaining zinc may be due to the incomplete washing out of zinc sulphate; but besides the sulphate there is always more or less zinc remaining unattacked by the acid, whether dust or shavings have been used. This is especially true when sulphuric acid is employed.

The causes of the failure of the acid to dissolve the zinc phenomenon seem to be:

1. The conditions of precipitation (strength of solution, metal- content, etc.) may be such as to form zinc alloys not readily attacked by the acid.
2. All commercial zinc contains more or less lead; and in some cases, lead is added intentionally (by immersing the shavings in a dilute solution of lead acetate) in order to promote precipitation by forming the lead-zinc couple. All the precipitates, therefore, contain more or less lead; and when they are treated with sulphuric acid the resulting insoluble lead sulphate tends to form a protective coating over the unattacked material, and thus prevent further solution of zinc and other impurities. Moreover, calcium is always present, and calcium sulphate has a similar tendency.

This residual zinc is certainly a disadvantage in the subsequent melting. Slags containing zinc usually carry high values in gold and silver; and, besides, more or less zinc finds its way into the bullion.

Mr. A. Whitby gives an analysis of acid-treated and calcined precipitates, said to illustrate the product of the Rand, S. A. R., which shows 7 per cent, of zinc oxide.

I have recently found 2.29 per cent, of zinc in a sample of acid-treated precipitates from a mill in the Black Hills, where the treatment is very carefully conducted.

Smelting with Lead

A great many of the early cyanide plants in the United States shipped their precipitates to smelters.

The method of refining at the Balbach Smelting and Refining Co.’s plant at Newark, N. J., is briefly outlined as follows :

The precipitates, tied up in paper sacks in parcels of from 1 to 5 lb., are charged from time to time upon the bath of molten lead in a cupelling-furnace. The gold and silver are quickly absorbed by the lead, and until the mass is well melted, the precaution is taken to keep all the drafts closed. Subsequently, the slag is removed by skimming.

After the cupellation the bullion is passed through the regular process of refining.

Sometimes the material is briquetted, and sometimes it receives an acid-treatment before cupellation; but this is not generally done, since the precipitates are charged in small parcels and the small amount of zinc introduced in this way does not materially affect the cupellation.

Smelting with lead has not been adopted by cyanide plants as a means of refining their product until quite recently. A plant in South Africa, which had been experimenting with it for some time, has lately adopted it. Mr. C. W. Merrill, metallurgist for the Homestake Mining Co., in 1900 worked out a system of refining in which smelting with lead was the prominent feature.

Tavener’s Method

Mr. P. S. Tavener describes the method employed by him at the Bonanza Limited, S. A. R., in August, 1899. This, I believe, is the only place in South Africa at which it has been used. He compares the operation to scorification on a large scale. It consists in smelting the fine precipitates and fine zinc in a reverberatory, with litharge, assay-slag, slag from former operations, silica and saw-dust (the latter stirred in with a rabble, after fusion, to reduce lead from the slag). The base bullion resulting from this operation is cupelled in a special cupelling-furnace.

Mr. Tavener claims that this process is less costly than acid-treatment; that it leaves no by-products (slags and litharge being worked over again); that it involves less loss in handling materials, and that it effects, in practice, a more complete extraction. It may be added, on the other hand, that his reverberatory is more expensive than ordinary melting-furnaces, and that the successful operation of the process requires considerable metallurgical skill.

Merrill’s Method

The method developed by Mr. Merrill at the Homestake plant, Lead, S. D., comprises a preliminary treatment with dilute hydrochloric acid; the removal of the liquid through a filter-press; a subsequent treatment with sulphuric acid; the washing and drying of the precipitates; mixing with litharge, borax, silica and powdered coke; sprinkling with lead acetate; and briquetting under a pressure of from 2 to 3 tons per sq. in. The briquettes are melted upon the hearth of an ordinary cupelling-furnace, the resultant slag is run off, air is turned on and the lead is cupelled in the same furnace. The resultant metal, 975 to 985 fine, is run into bars. Cupel-slags and bottoms are put through a blast-furnace, recovering the lead for the next cupellation; and the litharge is added to the next charge of precipitates. Mr. Merrill says that the blast-furnace slag carries less than $5 per ton in value; that the total cost of refining is less than 0.75 per cent, of the values contained in the precipitates, and that the loss (less than 0.1 per cent.) goes to the next precipitate. The total loss in refining is given as less than 0.1 per cent.

Comparison of Costs of Acid Refining & Lead Smelting

Table I. gives Mr. Tavener’s comparison of costs, and bullion recovered by the two methods, as practiced at the Crown Deep, South Africa. The clean-up was halved, one-half being treated by the usual acid-refining and the other half by lead-smelting, as described by Mr. Tavener.

For convenience, I have reduced pounds sterling to dollars, in reproducing Mr. Tavener’s figures.

“In the above, labor costs are not included. Labor was less in the lead-smelting.”

“The lead loss was estimated to be 12 per cent.; this, as well as the cost of material and labor for tests, is not given in the above, which will likely increase the cost of refining to from $0.04 to $0.05 per fine oz. of gold produced.”

“In three comparative tests, made upon the same material, lead-smelting yielded from 10 to 11 per cent. more bullion than did acid-refining.”

C. W. Merrill gives the cost of refining at the Homestake as $0.15 per fine oz. of bullion. Mr. Merrill has since furnished me with the very complete cost-sheet, given in Table II. It is unfortunate that the South African costs are not more fully given. A comparison would be most interesting.

The figures of the cost of acid-refining, given in Table III were obtained by me at the Horseshoe mill, Terry, S. D.

At this plant, clean-ups are made twice a month, which, together with the fact that the precipitation is carried on in 160 separate precipitation-boxes, makes the labor-item high.

The mill crushes wet, so that a large volume of solution is passed through the zinc in proportion to the tonnage of ore treated.

This seems to allow a large amount of lime, alumina, magnesia, etc., to separate in the zinc-boxes, greatly increasing the bulk of the precipitates and causing a high consumption of acid and fluxes.

The figures of cost include all items of expense from the time the washing of the zinc is commenced until the bullion is turned out in bars.

Mr. Tavener does not think a preliminary acid-treatment necessary; but Mr. D. J. Williams, in the discussion of Mr. Tavener’s paper, argues that the precipitates should receive acid-treatment before being smelted with lead. He gives the following objections to smelting when zinc is present:

(a) A great deal of experience is required to handle zinc in a furnace.
(b) this process is not one of oxidation, but the zinc must be fluxed. If oxidation sets in then, I maintain there will be a loss of gold and silver.
(c) Before all the zinc is fluxed, I should imagine that some would float on the surface of the molten mass, and carry by far the greater part of the gold and silver, as in Parkes’ process of de-silvering and de-aurizing lead bullion. If the zinc is allowed to oxidize or burn, undoubtedly there will be a great loss by volatilization.”

Various other arguments have been urged for acid-treatment prior to lead smelting.

Remove Zinc by Distillation

General Conclusions

Of the various methods of refining the gold-silver-zinc precipitates of the cyanide process, smelting with lead, after removal of the zinc, is the best.

Gold and silver are precipitated from cyanide solutions by zinc, as alloys or intimate mixtures of those metals with zinc.

It is possible to remove zinc more completely, and probably at a less cost, by distillation than by acid-treatment.

It seems possible to recover most of the zinc in a form at once available for use as a precipitant.

Outline of a Proposed Method of Treatment

Precipitates to be dried, mixed with lead and charcoal, and retorted. If mercury be present, it can be recovered at the outset of the process; later, zinc is condensed as dust, suitable for precipitation. The value of the zinc recovered would probably pay for the fuel used in retorting.

The residues containing the gold and silver to be smelted with lead in a furnace like the ordinary cupellation-furnace, iron-ore and silica being added in such proportions as to give a fluid slag, from which the lead separates. This is skimmed off and the lead cupelled in the usual way, the resulting litharge being used for the next operation.

The resulting bullion, after re-smelting, to be cast into fine bars of gold and silver which need no further refining aside from parting.