In the Estimation of Gold and Silver in Mercury and Amalgam, and the Preparation of Pure Mercury one can judge these subjects are so intimately connected that they are best treated together. It is a common delusion that only a negligible amount of gold or silver passes with the mercury through leather or cloth used in straining amalgam, and a still commoner one that mercury is completely freed from gold by distillation or “retorting.”
At the ordinary room temperature mercury dissolves at least 0.03 per cent, and up to 0.05 per cent, of gold, and I have found about 0.02 per cent, gold and half as much silver in squeezed mercury which had deposited most of its contained amalgam as crystals after long standing at a temperature just below the freezing point of water. The usually quoted value of 0.03 per cent, or 300 parts per million is equivalent to a good $ per ton. The actual content is often much higher, owing either to minute particles of suspended amalgam passing through holes in the straining cloth, or to straining at a higher temperature and retention of the dissolved amalgam, apparently as a supersaturated solution.
If the attempt is made, therefore, to separate the collected gold from the mercury in the amalgamation assay by squeezing through buckskin, it is easy to see that the results are entirely misleading. If we use only 5 g. of mercury with 100 g. of ore (or 3/4 oz. per pound) at least a few $ may be extracted per ton of ore and pass through the leather with the mercury. This method is entirely inadmissible except for the roughest estimation of coarse gold.
Even when the method of assaying the tailing is adopted, the use of squeezed mercury may lead to serious error; the retention of 0.5gr in the floured condition in 100 g. of ore making the tailing too high by at least a few $/ton.
In retorting mill amalgam an appreciable, though small, amount of gold is carried over with the quicksilver, probably owing to spurting or violent boiling. Estimates are that under the most favorable working conditions at least 5 parts of gold pass over per million of mercury condensed, so that the mercury carries at least a few $ worth of gold per ton. Often much more may be found in it. If mercury carrying $3 per ton is used in tests, at the rate of 20 g. per 100 g. of ore, the assay of the mercury will indicate a recovery which is too high by 60c. per ton of ore. If the extraction is calculated on the assay of the tailing the error thus introduced is unappreciable unless flouring is excessive.
By a cautious slow redistillation of this nearly pure quicksilver in a clean new iron retort, with a condenser of new iron pipe, it is obtainable practically free from gold. This is the best method to adopt where practicable. Vacuum apparatus may also be used for effecting the distillation under reduced pressure at a much lower temperature than is otherwise practicable, or glass condensers may be used, but neither is of much advantage for the present purpose. Hulett has shown that, if distillation is slow enough to avoid actual boiling, even such volatile metals as zinc and cadmium are carried over in only minute amounts.
It is still easier, at mills where quicksilver is bought in large quantities and comes directly from the mines, to test several of the best looking flasks for gold, and select the best found. Sometimes scarcely a trace is present in new quicksilver; occasionally a flask is found containing a considerable proportion of gold, possibly owing to the use of old mill flasks.
Tor the preparation of small amounts of gold-free mercury Darrow recommends treating a quantity with an insufficiency of nitric acid, so as to leave the gold in the undissolved portion. The filtered solution of nitrate is then precipitated by means of pure copper, such as electric wire. This is only suited to small amounts. Electrolysis of the nitrate is not practicable, as even the densest graphite electrodes are rapidly disintegrated. A simple method of removing gold from mercury was described and puts a thin layer in a shallow dish and covers it with a 2%, solution of potassium cyanide, adding a little sodium peroxide at intervals. The water requires renewal occasionally. I have found this effective but rather slow; stoneware or agateware pans may be used, but are somewhat attacked by the alkali.
In estimating the precious metal in small amounts of amalgam or in mercury, two methods naturally suggest themselves. One is to distill off the mercury, the other to remove it with nitric acid.
In practice it is found that the first invariably causes a certain loss of precious metal, while in acid parting the silver is always too low. It is recommended that the acid method be adhered to, so far as gold is concerned.
In case it is desirable to estimate silver also, it is recommended to take two portions of the material. In one the mercury is removed by nitric acid, finishing with hot and fairly concentrated acid; the remaining gold being then cautiously dried and heated to expel any residual mercury, inquarted, and parted as usual. The second portion is put into a small scorifier, covered with another inverted scorifier, and set in a comparatively cool muffle, which is then slowly heated to strong redness, taking care that there is a strong draft to carry the fumes outside. The residue in the scorifier is then covered with a layer of test lead, which is scorified a few minutes and cupelled. The scorifier used as a cover must be carefully examined for traces of adhering metal, and in exact work it is preferable to also scorify a little lead in the cover and add it to the main lot. Small lots of clean amalgam may be similarly treated in a covered cupel to avoid scorifying. From the total gold obtained by the wet method, and the ratio of gold to silver indicated by the fire assay, the amount of silver may be accurately determined; such refinements as the correction for cupel absorption being introduced if warranted.
A satisfactory method of recovering both silver and gold from amalgam or mercury is to heat it in a crucible under a heavy cover of assay flux, which is then fused and the resulting lead button cupelled. Thus one can recommend a layer of 60 g. of litharge, mixed with enough reducer to yield 25 or 30 g. of lead, and a cover of borax. Ordinary assay flux, with a little silica added, appears to work well. In any case extreme care must be taken to guard against the escape of mercury fumes into the laboratory.
For rapidly obtaining the gold from a small amount of amalgam the most convenient plan is to heat it in a miniature retort, made by closing a piece of glass tubing at one end and bending it to a slightly acute angle. The retort is very gradually heated to full redness and then quenched with water.
The device of placing it in a hollowed potato, or on an iron plate covered with half a potato, and heating till the potato is charred, is fairly satisfactory if the amount of gold is not too small.
When solid amalgam is treated with nitric acid a considerable amount of mercury is retained, even after long heating with strong acid, the gold protecting the mercury as it does silver. In the ease of mercury containing small proportions of dissolved gold heating with moderately strong acid leaves the gold nearly pure, so that the dried gold can be safely heated without spurting and the last trace of mercury thus removed.
The vapor of mercury has so high a density that a moderate draft is insufficient to carry it off, and heating in a muffle without proper precaution may lead to serious consequences. In regard to the danger of vaporizing mercury without using a condenser, the warning given by metallurgists may be appropriately quoted: “ Take heed lest the smoke or vapour go not into thy Belly, because it is a poisonous and cold Vapour, which amcth and killeth: for, he will find that it will there congeal and afterwards spoil his body.”
It is scarcely necessary to add that the cupellation of amalgam directly, or of gold containing more than a trace of mercury, causes the “spitting” of the cupelling lead, and the contamination of other alloys.
Extremely sloppy or dilute amalgam is very difficult to sample, as it is impossible to secure a uniform mixture of the solid and liquid portions. In sampling such material it is advisable to strain it through cloth at a rather low temperature, and weigh both solid and liquid portions. These can be assayed separately and the original content calculated, or similar aliquots can be taken from each portion and mixed for assay.