Table of Contents
While the reducing action of organic matter, of ferrous sulphate, and of hydrogen sulphide has frequently been invoked to account for the deposition of native gold and silver from ore-forming solutions, the high efficiency in this respect of certain of the metallic minerals which form the ore itself has not been sufficiently recognized. If a little finely powdered chalcocite is placed in a test-tube and 2 or 3 cc. of a dilute solution of gold chloride are poured over it, two or three shakes of the tube suffice to remove all the gold from the solution. The color of the solution changes from yellow to pale green, showing that copper is dissolved simultaneously with the precipitation of metallic gold.
If a piece of chalcocite is hung in a dilute aqueous solution of silver sulphate (1/40 normal), a coating of silver begins immediately to form on the mineral, and within a few moments a beautiful silver tree has developed, similar to that formed when metallic zinc is immersed in a similar solution.
When similar experiments are conducted with other sulphides and with certain arsenides and sulph-arsenides, especially those known to be commonly associated with native silver in ore deposits, it is found that these minerals differ greatly in their efficiency as precipitants, especially in their effect on silver solutions. The tests made with most of the minerals were qualitative only, but the reactions of a silver sulphate solution on chalcocite and niccolite were submitted to quantitative study to determine the nature of the reactions involved. Quantitative studies of the reactions with other minerals are now in progress in the laboratory of the U. S. Geological Survey; but as the phenomena seem to have an important application in the science of ore deposits, particularly in secondary enrichment, it has been thought best to outline the field of investigation in this preliminary paper and to make the main results immediately available to economic geologists.
The earliest important work on the precipitation of gold and silver by metallic minerals was that of W. Skey, printed in 1871 in a New Zealand publication. His work appears until recently to have escaped the attention of economic geologists; it blocked out, however, one of the most interesting and promising fields of research in the science of ore deposits.
Another line of investigation inaugurated by Skey involves the measurement of the electrical potentials developed when two metallic minerals are connected as poles of a galvanic circuit and immersed in some conducting liquid which does not react actively with them. This line of research has recently been followed by Gottschalk and Buehler and E. C. Wells.
These writers have arranged the commoner metallic minerals in the order of the electrical potentials developed when they are in contact with a conducting liquid, and Wells concludes that “ such potentials simply measure in an electrical way the oxidizing power of the solutions, the mineral playing a very insignificant part. The potentials in metallic salts solutions depend entirely upon the nature and concentrations of the solutions.”
It is now recognized that most reactions of oxidation and reduction may be effected by the use of the electric current in place of chemical agents, and, conversely, that an electric current may readily be produced by a proper arrangement of the components of any one of these reactions. As a consequence, by measuring the relative strengths of the currents developed when various sulphides are successively connected as parts of a galvanic circuit, it is possible to obtain a measure of the relative rate at which oxidation or reduction takes place. Such methods of study may prove useful in determining the relative efficiency of various minerals in reactions of the type described in this paper, but careful analytical studies are essential to their apprehension and for the present appear to form a more profitable line of research.
Experiments With Silver Sulphate Solution
In these experiments the minerals were selected so as to be as free as possible from admixtures of other minerals. Considerable variation is known to exist in the composition of natural minerals referred to the same species, and such variations may influence greatly their efficiency as precipitants of the precious metals. As a matter of record, therefore, data are given in a foot-note, showing the source of the material used. The experiments were conducted at ordinary temperatures. Pieces about 0.25 cc. in size were taken and were immersed in small evaporating dishes in 20 cc. of 1/40 normal solution of silver sulphate. The solution was neutral at the beginning of the tests. The progress of the reaction was observed at appropriate intervals.
Silver sulphate was chosen as the reagent in these experiments because it is well known that the solutions of the oxidized zone in sulphide ore deposits are sulphate solutions and also because the sulphate is much more soluble than either the chloride or the carbonate, the relative solubility in water at 25° C. being approximately as follows :
Silver chloride…………………………………………………….0.002
Silver carbonate………………………………………………….0.033
Silver sulphate……………………………………………………..8.01
The experiments reveal notable differences in the efficiency of the sulphides, arsenides, and sulph-arsenides treated as agents for the reduction of metallic silver from dilute aqueous solutions of its sulphate, and various reaction products in addition to metallic silver are developed in most cases.
In tentatively grouping the minerals tested according to their efficiency it is recognized that the speed of the reaction could only be roughly estimated by the methods adopted and that some of the minerals carried impurities which may have had an important influence on their activity.
Very strong: Chalcocite and niccolite.
Strong: Covellite, enargite, bornite, tennantite, alabandite, and possibly cobaltite.
Moderate: Smaltite, marcasite, pyrrhotite, chalcopyrite, and arsenopyrite.
Weak or inactive: Cinnabar, stibnite, pyrite, galena, millerite, sphalerite, jamesonite, orpiment, and realgar.
Experiments with other Silver Solutions
The silver sulphate solution used in the experiments just described was neutral. In order to determine the effect of acidity upon the reactions, a solution of silver sulphate was rendered acid by the addition of sodium acid sulphate (NaHSO4) and its action on chalcocite, chalcopyrite, and niccolite was observed. No marked difference was noted in the speed of action of the acid and neutral solutions. A minor feature of interest was the development of delicate feathery silver crystals on the chalcocite immersed in the acid solution, while that immersed in the neutral solution formed brilliant spangles.
In order to approach more closely the conditions actually found in certain ore deposits a solution was made up which except for its silver content corresponded in composition to the vadose water collected by Weed from the second level of the Mountain View mine, at Butte. This water was dripping from the roof of the drift and was depositing chalcanthite. The artificial solution had the following composition :
CuSO4. 5 H2O = 180 g. per liter.
ZnS04. 7 H2O = 2 g. per liter.
Ag2SO4 = less than 4 g. per liter.
Chalcocite, bornite, covellite, galena, chalcopyrite, and niccolite were treated with this solution and silver appeared to be precipitated as rapidly as from the neutral silver sulphate solution. With the neutral silver sulphate solution galena precipitated only very minute amounts of silver after many days’ immersion. With the cupriferous silver solution the action was more rapid, though still slow as compared with the other minerals tested. In general, the presence of copper in the solution appeared to facilitate rather than to retard the silver precipitation.
A simple experiment with solutions of silver benzol sulphonate was performed to demonstrate the role of hydrolytic action in reactions of the kind under investigation. This organic salt of silver is very soluble in water; if chalcocite is immersed in a saturated solution of the salt, silver is precipitated only with extreme slowness; if, however, the solution is diluted with about 30 volumes of water silver begins to deposit rapidly and the reaction continues until silver has been completely removed from solution. In the equation given on page 231, representing the reaction of silver sulphate solution on chalcocite, no account is taken of the part which the elements of water play in the changes involved. That water does enter into the reactions is evidenced by this experiment.
Quantitative Experiments
Niccolite and Silver Sulphate Solution
The so-called niccolite used in these experiments has the composition :
Weighed quantities of the finely powdered mineral were digested in dilute silver sulphate solutions containing known amounts of silver sulphate. Action began promptly. Free silver was deposited and the solution became green. After the lapse of several hours the gray residual mineral together with the silver deposit was collected on a filter. After deducting the weight of the silver the mineral residue was found to represent 35.12 per cent, of the niccolite treated. It consisted of cobalt, iron, arsenic, and sulphur, distributed as follows:
The composition of this unattacked residue obtained from the so-called niccolite corresponds to the formula for cobaltite, namely Co(As(S))2.
The weight of silver deposited with the residual cobaltite amounted to 260.1 per cent, of the total weight of the mineral tested.
The portion of the mineral dissolved by the silver sulphate solution consisted only of nickel and arsenic. The nickel amounted to 28.26 per cent, and the arsenic to 36.50 per cent, of the original mineral. Since these results correspond to one atom of arsenic for every atom of nickel dissolved, it is apparent that the material attacked by the silver sulphate solution is true niccolite (NiAs). The arsenic was present as arsenious acid. Since the dissolved nickel, dissolved arsenic, and deposited silver are proportionally, nickel, one atom; arsenic, one atom, and silver, five atoms, the reaction between true niccolite (NiAs) and dilute silver sulphate solution may be expressed by the equation
2 NiAs + 5 Ag2SO4 + 3 H2O = 2 NiSO4 + As2O3 + 3 H2SO4 + 10 Ag.
From these considerations the so-called niccolite may be regarded as consisting of 64.76 per cent, of true niccolite and 35.12 per cent, of true cobaltite. Microscopic study of a polished piece confirmed the conclusion that it was a mixture of two minerals. That silver sulphate solution may be used advantageously to throw light on the nature of certain complex minerals is apparent, for by its agency two distinct mineral species, cobaltite and niccolite, have been quantitatively parted. In the original specimen, which contains both cobalt and nickel, apparently as inseparable companions, it is noteworthy that the silver sulphate solution assigns all the cobalt to the sulpharsenide, cobaltite, and places all the nickel in the simple arsenide, niccolite. The pyritous structure of true cobaltite may account for its failure to yield to the oxidizing action of dilute silver sulphate solution.
Chalcocite and Silver Sulphate Solution
That free silver is deposited if metallic sulphides are digested in a silver nitrate solution has been mentioned by several observers, but their statements concerning the products formed lack unanimity. For instance, Heumann observes that from artificial cuprous sulphide (Cu2S) free silver is deposited along with a gray powder, which he considers to he silver sulphide. Schneider corroborates Heumann. S. Meunier states that he observed free sulphur among the products of the action of silver and gold solutions on several metallic sulphides.
In view of these divergent statements concerning the action of sulphides on silver solutions, it seemed advisable to determine the action of chalcocite on dilute silver sulphate solution. In no case was free sulphur formed, but the sulphur of the chalcocite was retained and held in combination as silver sulphide.
Total Ag deposited = 255.14 per cent, of chalcocite decomposed.
S in residue = 18.34 per cent, of chalcocite decomposed.
Cu dissolved = 75.81 per cent, of chalcocite decomposed.
These figures correspond to: sulphur, one atom; copper, two atoms; and silver, four atoms, and the reaction may be expressed by the equation
Cu2S + 2 Ag2SO4 = 2 CuSO4 + Ag2S + 2 Ag
While water does not appear in this equation it is clear from the experiment already cited on page 228 that it plays an essential part in the reaction. In a dilute silver acetate solution chalcocite acts in like manner.
Covellite containing 66.06 per cent, of copper and 33.87 per cent, of sulphur, immersed for four days in a dilute silver sulphate solution, yielded only 54.97 per cent, of its copper. This loss of 83.2 per cent, of copper and the deposition of an equivalent amount of silver as free silver and silver sulphide suggest that a more intimate structural relation exists between chalcocite and covellite than is indicated by the conventional formulas, Cu2S and CuS, which are used respectively to represent the molecular masses of these two minerals. The study of the constitution of covellite is now in progress.
Experiments with Gold Chloride
The minerals used were from the same localities as those used in the tests with silver. The experiments were conducted at ordinary temperatures. Pieces about 0.25 cc. in size were taken, and were immersed in small evaporating dishes in 20 cc. of gold chloride solution carrying about 7 g. of gold per liter. The solution was slightly acid to start with. The progress of the reactions was observed at appropriate intervals.
The gold precipitated by the action of certain of the minerals was mainly dark brown and porous, and the reaction continued until all the gold of the solution had become exhausted. The gold developed on other minerals, on the contrary, was yellow and compact, and when the piece was wholly plated this coating seemed to protect the mineral from further action. For this reason it is not possible to classify the minerals according to their activity in precipitating gold as closely as was possible in the case of silver. Under natural conditions the sulphides would seldom become completely coated with gold and the inhibitory action due to such complete plating would not be felt.
All of the metallic minerals tested that were efficient as precipitants of silver were also efficient as precipitants of gold.
Galena, pyrite, stibnite, and millerite, which were inefficient as precipitants of silver, were active as precipitants of gold.
Sphalerite and cinnabar acted only weakly in precipitating gold.
Effects of Thin Coatings of Cuprous Sulphides
The occurrence of very thin films of gray secondary chalcocite on chalcopyrite is a common phenomenon in the upper portions of many ore bodies. A specimen from Gilpin county, Colo., showing such films was treated with silver sulphate solutions. The chalcopyrite precipitated silver only with extreme slowness but the thin chalcocite coating was a very efficient precipitant. The reaction forms, in fact, a convenient means of identifying such gray coatings as probably chalcocite when they are too thin to be detached and tested.
Peacock-colored coatings are also common on the chalcopyrite of some veins and such coatings may be readily produced artificially by placing a piece of chalcopyrite for a few moments in contact with an iron nail in a solution of copper sulphate. A piece so tarnished will precipitate silver and gold with great readiness, whereas the uncoated chalcopyrite acts only very slowly. It is clear, therefore, that thin films of cuprous sulphides on chalcopyrite or other minerals may be almost as efficient as larger bodies of such sulphides in precipitating the precious metals.
Application in Ore Deposits
Association of Native Silver with Copper Ores
The results recorded above find their principal counterpart in nature in the phenomena of secondary enrichment in ore deposits. The subject is a large one, and it is only necessary here to cite a few examples to show the influence of such phenomena upon the distribution of values in ore deposits.
If chalcocite exercises such a strong precipitative action upon the precious metals when these are carried in neutral or acid solutions, we should expect that the precious metals would be re-precipitated as soon as they reach the upper part of the chal- eocite zone. Although most of the great copper deposits of the world in which chalcocitization has been observed carry only low values in gold and silver in the original ore, such evidence as is available indicates that the chalcocite has exerted a precipitative influence upon the precious metals.
In the copper veins of Butte, Mont., there are present nearly all of the copper minerals which these experiments have shown to exert a strong precipitative action on silver and gold. Here the conditions should have been almost ideal for testing the efficiency of such action under natural conditions. Unfortunately, the precious metal content of the original ores is very low and little detailed information is available in regard to the occurrence of precious metals in the upper portions of the copper veins. In the Butte folio appears the following statement: “ Native silver has been found in both the copper and the silver mines. It is the only silver mineral that it has been possible to recognize in the copper veins,” a statement that suggests the efficiency of the copper minerals in reducing silver salts in solution to the metallic state.
According to Weed, “ Native silver is common in the silver veins and also in the oxidized portions of the copper veins. It has been observed in mossy aggregates and coatings, in fracture planes in glance [chalcocite] ores, and in bornite. Wire silver and fibrous silver also occur in cavities in bornite and glance [chalcocite], particularly in the upper levels of the Parrot mine.”
Another line of evidence consists in a comparison of the average silver values in the upper and lower portions of certain veins. Weed states that for the upper levels of the Gagnon mine, at Butte, the ore carried an average of 3 oz. of silver to each 1 per cent, of copper, whereas the average silver content of the ores extracted from lower levels in 1905-06 was only half as great: namely, 1.5 oz. of silver to each per cent, of copper.
It appears probable, therefore, that the precipitative action of chalcocite and other sulphides of copper played a part in the localization of the silver values in the upper portions of the Butte copper veins.
The intimate association of bornite, chalcocite, and native silver has been noted by the writer in the Up-to-Date mine near Caribou, Boulder county, Colo. The rich ores of this mine carry chalcocite, bornite, native silver, and calcite (non-manganiferous) as their principal minerals, with subordinate amounts of covellite. The ores of this type thus far obtained were found within from 50 to 100 ft. of the surface and occur as small veinlets in and as replacements of altered pyroxenite. Some of the veins are an inch or so in width, and one of them yielded a specimen of native silver 6 by 8 in. by 0.25 to 1 in. thick. From this size they grade downward to exceedingly minute veinlets less than 1 mm. wide. The irregular form of the smaller veinlets shows that they probably developed by metasomatic replacement of the rock proceeding outward from a very minute fracture. The irregular mode of association of the native silver and the sulphides indicates that they were contemporaneous crystallizations. Covellite when it occurs is intergrown with bornite or chalcocite, bears no relation to vugs or cracks and evidently crystallized with the other sulphides.
There is little doubt that the rich silver-copper veins of this mine are the result of downward sulphide enrichment, the materials being supplied by primary veins carrying galena, sphalerite, chalcopyrite, and calcite, and other veins carrying chalcopyrite, pyrite, and quartz which are also exposed in these workings. The detailed evidence of such origin is not germane to the present discussion. In most of the natural occurrences previously described the native silver was deposited distinctly later than the sulphides, simulating closely the conditions of the experiments which are here reported. In the present case the copper sulphides and the native silver were precipitated together. The chemistry of the process as regards silver precipitation is believed to be essentially the same in either case.
Occurrences similar to that in the Up-to-Date mine are not uncommon. Beck described a similar association of minerals in the copper shales of Germany. He says: “ In addition to the small grains [of sulphides] there also occur in the Kupferschiefer fine bands, mostly parallel to the bedding, of bornite and chalcocite, and along bedding-planes and cross-fractures, coatings of chalcocite, bornite, chalcopyrite, and native silver.”
These instances, while by no means exhausting the literature, suffice to show that the association of silver with chalcocite, bornite, and related copper minerals has been frequently noted. Doubtless the presence of native silver in intergrowth with such copper minerals will be found to be a common phenomenon as more copper ores are studied microscopically in the polished sections.
Association of Native Silver with Cohalt and Nickel Arsenides
The experiments showed that niccolite, and, to a lesser degree, so-called cobaltite and smaltite, were comparable with chalcopyrite, bornite, and covellite in the strength of their reducing action upon an aqueous solution of silver sulphate.
The frequent association of native silver with cobalt and nickel arsenides is too well known to require lengthy comment. It is perhaps sufficient to cite as an example the Cobalt district of Ontario. Dr. S. F. Emmons, in describing this district, says:
“The more common minerals of the rich vein deposits are the arsenides of cobalt and nickel, smaltite with some chloanthite, cobaltite, and niccolite, associated with native silver. Less frequent are native bismuth, the silver minerals pyrargyrite, proustite, dyscrasite, and argentite, the nickel sulphide, millerite, with occasional mispickel and tetrahedrite. The ordinary sulphides pyrite, galena, and zincblende are occasionally found in the wall rocks, but apparently do not form an essential part of the deposits. The gangue minerals are calcite, with a little quartz, but both are in relatively subordinate amount in the rich parts of the veins.”
“ It is generally recognized that the native silver is of distinctly later deposition than the cobalt-nickel arsenides—indeed, the evidence in the mines is most conclusive. Very often minute cracks may be seen crossing the cobalt mineral and gangue rock which are filled with films of native silver, and the flake or sheet form in which the silver is so often found shows that it has grown in such cracks.”
In the experiments conducted with niccolite and so-called cobaltite and smaltite, one or more reaction products, in addition to native silver, were abundantly developed. Although the nature of these other products has not been determined, it is suspected that some of them are silver salts, and it is noteworthy, therefore, that several other silver minerals besides native silver are present in the cobalt veins.
Associations of Native Gold with Sulphides
“Without touching the question of the apparent preference manifested by gold for certain sulphides in deposits that are clearly primary, it is possible to cite numerous instances where the sulphides seem to have exerted a reducing influence during the process of secondary enrichment.
Reno Sales has found distinct and fairly good-sized particles of native gold on the surface of chalcocite crystals from the 1,000-ft. level of the Leonard mine, at Butte.
In describing secondary enrichment in the pyritic deposits of Huelva, Spain, Finlayson says:
“The lower limit of the gossan is sharp and well-defined, and the line of contact between gossan and sulphide ore is sometimes marked by an earthy zone carrying considerable values in gold and silver. This has been described by J. H. L. Vogt, who has pointed out that it represents a concentration, during secondary enrichment, of the traces of gold and silver in the original ore, the precious metals dissolved in ferric sulphate being precipitated by the reducing influence of the pyrite.”
How Metallic Minerals Affect Silver and Gold Precipitation
The preliminary experiments which are here described show that certain sulphides, arsenides, and sulph-arsenides of copper, nickel, and cobalt precipitate metallic silver very efficiently from dilute aqueous solutions of silver sulphate. As the waters descending through the upper portions of most sulphide ore bodies are known to be sulphate waters, similar precipitative actions would be expected under certain natural conditions. The frequent association of silver in ore deposits with chalcocite and bornite, and particularly with niccolite and cobaltite, minerals which in these experiments were among the most efficient precipitants, warrants the belief that such reactions are of importance in secondary enrichment of ore bodies.
The more common sulphides, such as pyrite, galena, and sphalerite, were relatively inactive as precipitants of silver from aqueous solutions of its sulphate.
The quantitative results obtained with niccolite and chalcocite indicate that the essential chemical changes in reactions of this type are due to oxidation through the hydrolytic action of water. It is apparent, therefore, that certain water solutions may act as potent oxidizing agents below the ground-water level.
Unlike gold, silver forms a large number of natural compounds. It is probable that the presence of certain other substances in the vein solutions, notably arsenic and antimony, may so modify the reactions that silver is precipitated not as native metal but as a compound such as polybasite or proustite. The conditions under which such compounds are formed forms an allied and very important field of research.
The experiments indicated that nearly all of the sulphides and arsenides common in ore deposits were capable of reducing gold from a solution of its chloride, although important differences in the rapidity of the precipitation were observed with different minerals. Most of the minerals that are especially efficient as precipitants of silver are also effective precipitants of gold, and a number of other minerals such as galena, pyrite, stibnite, and millerite that are inefficient in precipitating silver are efficient in depositing gold.
It is known that the waters descending through the upper portions of sulphide ore bodies universally carry chlorides, and it is probable that these chlorides have effected the solution of the gold. It is probable, therefore, that phenomena similar to those exhibited in the experiments with gold chloride solution play an important part in secondary enrichment in gold.
While the phenomena here described find their most immediate application in secondary enrichment, it is perfectly possible that such reducing effects of the sulphides may be responsible in part for the primary association of the precious metals with certain sulphides in preference to others. It is recognized, of course, that certain mineral associations in ore deposits are probably the result of processes analogous to differentiation in rock magmas, but it is quite possible that other associations, such as the apparent preference of gold for chalcopyrite and tetrahedrite rather than for pyrite in deposits carrying these three minerals, may be due to differences in the reducing power of these sulphides themselves. Light could probably be thrown upon this point by the investigation of the reducing effect of various sulphides upon silver and gold salts dissolved in solutions having the composition of certain deep mine waters. The effect of increase of temperature on the reactions should also be studied.
It is a generally recognized fact that the purity of alluvial gold is greater than that of the veins in the neighborhood. This superiority in fineness has generally been explained by the well-known fact that silver is more readily soluble in natural waters than gold, and is by them removed from the natural alloy, thus increasing its purity. Mr. Lindgren has recently discussed this matter at some length in a report on the Tertiary gravels of California, and has presented a large number of statistical data leading to the same conclusion. It has been thought by certain geologists that this refinement of the gold was accomplished by solutions circulating through the gravels themselves, but Mr. Lindgren states that “ so far as the Tertiary gravels of California are concerned, the conclusion of the writer is that solution and precipitation of gold have played an absolutely insignificant part.” Under the conditions of the experiments here reported it was found that nearly all of the metallic minerals common in precious metal deposits were capable of precipitating gold, while a much smaller number, and these not the most common ones, were active precipitants of silver. When it is remembered that the source of the placer gold is the oxidized zone of the original deposit, and that the gold may have been dissolved and redeposited several times within the vein before erosion carried it into the alluvium, it seems probable that such selective precipitation maybe a factor in this natural refining of gold.
