Acid Leaching

The experimental work on the oxidized copper ore at the New Cornelia mine at Ajo, Ariz., ended on Jan. 12, 1916. On that date final decision was made on the general nature of the process to be used in the 5,000-ton leaching plant, and on many of the details, as far as experience on a 40-ton scale could decide them.

With the approval of the Board of Directors and the General Manager, John C. Greenway, we have compiled what seem to be the most interesting data on the results obtained during the experimental period. The most important part of the data resulted from the operation of a 1-ton and a 40-ton plant at Ajo.

Experimental work on Cornelia oxidized ore dates back to April, 1912, and has been going on nearly continuously since that time. A good many variations from the original idea have been tried, but the process finally decided upon is in principle and in all of its details a simple one.

The history of the leaching work on this ore has been brought down to about a year ago in papers presented to this Institute by Stuart Croasdale and Dr. L. D. Ricketts. The present paper will, therefore, deal principally with the results obtained during the last year of the work.

Preliminary Tests by Mr. Croasdale

Preliminary tests were begun by Stuart Croasdale in July, 1912. A number of important points were definitely decided by his work. Among these were:

1. New Cornelia oxidized ore, not very finely crushed (to about 2- to 3-mesh) showed good extraction with 5 per cent, sulphuric acid. It appeared that better than 80 per cent, extraction could be expected.
2. The consumption of acid for this extraction was not prohibitive. It was apparently 3.5 to 4 lb. gross per pound of copper recovered.
3. It was proposed to make cement copper. The iron consumption per pound of copper recovered was a little over 1 lb.
4. A column 12 ft. deep of ore crushed to this fineness could be successfully percolated.

During Mr. Croasdale’s work the plan in view was a very simple one indeed. The ore was to be leached with sulphuric acid. Cement copper was to be produced by the use of metallic iron. As Dr. Ricketts explained to the Institute nearly 2 years ago, the oxidized Cornelia ore could apparently be leached and cement copper produced at a profit, if sulphuric acid and scrap iron were bought in the open market and shipped to Ajo; and cement copper shipped out, all solutions being thrown away after removal of the copper.

Mr. Croasdale showed also that the amount of substances other than copper, taken into solution under his conditions of crushing and leaching, is small. This naturally suggested the possibility of a closed leaching cycle with electrolytic deposition and at least partial regeneration of the acid required for solution of the copper. Early experiments showed that iron, which is dissolved from the ore to a certain extent, and oxidized at the anode during electrolysis, was a troublesome factor.

There seemed to be three possible ways of meeting,this difficulty:

1. The use of a diaphragm;
2. the use of a depolarizer;
3. purification from iron before electrolysis.

Antisell Process

About this time there came to the notice of the manager good results that were being obtained with a special anode, the invention of F. L. Antisell. This anode was in the form of a long, deep, narrow box with wood veneer sides, containing graphite electrodes and having the space between graphite and wood filled with selected coke.

A cyclic process was tried with these anodes. Ore crushed to 4- mesh was leached 3 days by upward percolation, with a practically constant solution containing 3 per cent. H2SO4, 2 per cent, ferrous iron, 0.5 per cent, ferric iron and 2 per cent. Al2O3. Sulphur dioxide gas was used to keep down the ferric iron concentration.

The process gave good results. The anode was, however, rather cumbersome and the absorption of the sulphur dioxide was not satisfactory, presumably owing to the rather high acid content of the solution.

The following is a summary of average results obtained by the Antisell Process: Tons of ore leached, 75; heads, 1.68 per cent. Cu; extraction, 77.6 per cent.; current density, 6.54 amp. per square foot; voltage, 1.23; production of copper, 1.72 lb. per kilowatt-hour; H2SO4 in solution, 2.73; sulphur per pound of copper, 1.06 lb.; total days run, 40.2; total cathodes produced, 803 lb.

Process of Pope and Hahn

While these tests were being carried on, another process was being developed at Raritan by F. A. Pope and A. W. Hahn. The Antisell idea was in a sense a combination of diaphragm and depolarizer, without any attempt to remove the iron. Pope and Hahn planned to remove the iron (and most of the other impurities as well) before sending solution to be electrolyzed.

Their process was cyclic. Ore crushed to 4-mesh was leached for 3 days with dilute sulphuric acid, the counter-current leach being so distributed as to permit of drawing off two classes of solution—one high in iron, the other low. The high iron solution was heated to about 200° F. with the addition of the requisite amount of finely ground (90 per cent, through 200-mesh) copper oxide in the form of roasted ore. The mixture of solution and oxides was agitated for 3 or 4 hr. and then put through a filter press. The result was the removal of nearly 90 per cent, of the total iron and 70 to 80 per cent, of the alumina from solution. An electrolyte high in copper and low in all impurities was produced and the precipitate, as filter cake, was granular and easy to wash and handle.

The other portion of the leaching solution, low in iron, was electrolyzed and returned to the cycle until its iron content was high enough to require purification.

Results of the Pope-Hahn process were very satisfactory, but the plant had hardly started when the war broke out and all experimental work was stopped. The following summary will indicate the results obtained: Ore leached, 329 tons, 10 charges; average copper heads, 1.12 per cent.; copper tails, 0.30 per cent.; water, tails, 9 per cent.; copper per kilowatt-hour, 1.0 lb. About 40 lb. of roasted Miami concentrate per ton of ore handled took care of impurities; 92 per cent, of the copper in the concentrate was recovered.

Working the Pope-Hahn process would have required the installation of roasting plant, fine-crushing plant, agitators and filter presses and the purchase of high-grade sulphide ore or concentrates. It offered decided advantages, among them a practically pure solution for electrolysis, together with the regeneration of a high percentage of the acid necessary for leaching.

Greenway Process

Before the study of the effect of impurities had progressed very far, an interesting fact was noticed. If neutral solution from the leaching system was circulated through fresh ore, a large part of the iron in solution was precipitated, presumably as a basic sulphate. This basic sulphate was apparently not readily soluble in a solution containing more acid. A possible method of purification from iron seemed to be indicated and the results of small-scale tests were so encouraging that a 1-ton plant was built. In the first tests, ore crushed to about ½ in. was leached for 6 days. The oldest charge of ore was in contact with fresh leach from the electrolytic cells. This leach carried about 3 per cent, of free H2SO4. Solution passed in counter-current through six tanks containing ore, fend the acid of the leach was neutralized by the time it reached the fourth or fifth tank. In the last two tanks a neutral solution was in contact with fresh ore and here the iron was deposited.

The results obtained will be evident from the data of Table 2. The ferric iron was under good control for more than 100 complete cycles.

At this time it was already practically settled:

1. That a good extraction could be obtained with dilute sulphuric acid.
2. That the acid consumption was not prohibitive.
3. That electrolytic copper could be produced successfully.

There remained, however, some important problems to be solved. Among these were:

1. Crushing; from two points of view: (1) for percolation in a deep bed; (2) for extraction.
2. Further study of the fouling of the solution, including the general accumulation of soluble salts, but especially those which might affect extraction or power efficiency in electrolysis.
3. Circulation and general handling on a small commercial scale.
4. Any factors other than those under (2), which might affect electrolysis itself or the quality of copper produced.

The 1-ton plant was therefore continued in operation and a plant to handle 40 tons of ore per day was built.

It is fortunate that this larger unit was built, for a good many things that had seemed easy to do in the 1-ton plant refused to work at all in the 40-ton. Fouling by iron and aluminum salts kept on increasing, the ferric-iron content of the solution rose to a point where the ampere efficiency in electrolysis was poor, and various undesirable phenomena appeared.

So the 40-ton plant was kept in continuous operation for nearly a year and all the factors that had been studied in the 1-ton plant were taken up, one at a time, and worked over again until finally it seemed that they could all be brought under complete control.

Final ACID Leaching Process

The final process decided upon was as follows: Ore crushed to about 4-mesh; leached 8 full days by counter-current; washed with three counter-current wash waters (or possibly four); solution, practically neutral during last 2 days of contact with ore, is sent through reduction towers, where it meets sulphur dioxide in counter-current. The ferric iron is thus reduced to below 0.4 per cent. Thence it goes through a revolving tumbler in contact with cement copper. There the ferric iron is still further reduced and a corresponding amount of cement copper passes into solution. From the cement-copper tumbler the solution passes to a settling pond and then into the electrolytic cells, where a part of the copper is removed; then back into the leaching system, passing first through the oldest ore which has already been leaching for 7 days, and so on to begin the cycle again.

Important ACID LEACHING Factors 

In a cyclic process such as this, no single factor or step can be said to be the most important. We can, however, consider a list of those factors that were studied with especial attention and can offer definite figures on the results to be expected.

Extraction

Both in the 1-ton and the 40-ton plants, it was easy to obtain 80 per cent, extraction, using about 3 per cent, sulphuric acid and 8 days of counter-current leaching by upward percolation on 4-mesh material. An extraction of 83 per cent, was reached over considerable periods of time. The continuous attainment of this figure was limited by other factors than the nature of the ore or the size to which it was crushed, such as the fouling of solution, with the result that salts carrying copper actually crystallized in the ore in the leaching tanks and were not completely dissolved by washing.

Incomplete washing was, of course, another reason for low extraction under otherwise normal conditions, and the possibility of improvement in this point is clearly shown by the wash-water data of Table 6.

Keeping well within the time limit for washing set by the fixed cycle, it should apparently be possible to get an extraction of 82 per cent, or better.

Acid Consumption

The 1-ton plant, over its whole life of 413 cycles and under all sorts of favorable and unfavorable conditions, showed a net consumption of 1.65 lb. of 100 per cent. H2SO4 per pound of copper leached. This was somewhat lower than the figures given by the preliminary tests on the ore.

The 40-ton plant apparently did not do so well. During its 301 cycles, the average net consumption was 2.8 lb. of acid per pound of copper leached. The difference in favor of the 1-ton plant is probably due to the large leakage and general losses in the 40-ton plant.

In designing the plant which is now being built at Douglas to furnish acid for Cornelia, 3 lb. has been taken as a conservative figure for acid consumption.

Power Consumption for Electrolysis

In the 1-ton plant, with lead anodes, the average was 0.934 lb. copper per kilowatt-hour at 8 amp. per square foot of cathode surface. This was for all the conditions, good and bad, under which the plant was operated. With graphite anodes, also under all sorts of conditions, the figure was 2 lb. copper per kilowatt-hour at 8,5 amp. per square foot.

In the 40-ton plant, with lead anodes, at an average current density of 6.9 amp. per square foot, the average production under all conditions was 0.87 lb. copper per kilowatt-hour. With graphite anodes, the figure was 1.65 lb. It should be said, however, that the ferric iron was never under proper control during these tests with graphite in the 40-ton plant. During series VI of the 40-ton plant, with current density 6.4 amp. per square foot, the copper yield was 1.04 lb. per kilowatt-hour. This run was with lead anodes and under fair conditions, so far as ferric iron was concerned. In the final run with lead anodes, and with both ferric iron and general fouling under still better control, an even higher figure was obtained.

Lead anodes will be used in the big plant. The reasons for this choice will be given later.

Control of Ferric Iron

Quite early in the study of the conditions for good electrolysis it became evident that the ferric-iron content of solutions must be kept within a low limit. The 1-ton plant took care of itself over a long period of time, as far as the ferric-iron content and the general fouling were concerned, without the use of sulphur dioxide or any other outside reagent. Iron and alumina were apparently precipitated on the ore.

The 40-ton plant started off in the same way. But before long the ferric iron began to go up and the ampere efficiency to go down. This is evident from the table showing general results on the 40-ton plant. Series II began with fresh acid and lasted 40 days. During this period 0.99 lb. Cu per kilowatt-hour was the electrolytic yield. The current efficiency during this period of low ferric iron and clear solution was 83.3 per cent. During period III the same solution was continued in use, but the ferric iron was not under control, and kept increasing. The analysis of the solution on July 22, in the middle of this run, is shown in column III of Table 4. The ferric iron content had risen rapidly from 0.180 per cent, on June 20 to 1.222 ger cent, on July 22. The result was that during period III, the current efficiency was only 70.8 per cent, and the yield was 0.85 lb. copper per kilowatt-hour. From this time until the

Column I gives solution analysis at beginning of a run, where fresh acid has been used, and before electrolysis has been started.
Column III is the analysis of solution when reduction of ferric iron was begun.
Column V shows solution at. maximum concentration. Cold nights caused the separation of a large amount of FeSO4, CuSO4, 24H2O.
Column VI is an analysis of solution under normal conditions, after “bleeding” regularly for about 40 days.

end of the experimental work, sulphur dioxide gas was used regularly to control the ferric iron. With lead anodes a low content of ferric iron is not necessary. The solution may contain 0.3 or 0.4 per cent, without any marked effect on the current efficiency.

Sulphur Consumption

To maintain satisfactory control of the ferric iron was, at this point in the work, the greatest apparent difficulty. To follow the course of the tests it will be necessary to keep in mind both Tables 3 and 4. The first of these shows the conditions under which the 40-ton plant was operating; the latter, the analysis of the solution at various times.

During period III, the acid electrolyte was circulated through a tower about 2 ft. square, in contact with sulphur dioxide. The acid electrolyte was taken from the electrolytic system and returned to the same system. The weight of sulphur that could be burnt and brought in contact with solution as SO2 was too small, and the ferric iron kept on increasing. Under these conditions, we were burning 1.5 lb. sulphur per pound of ferric iron reduced.

Series IV began with only a single change, but this proved to be a very important one. The neutral advance from the leaching system was circulated through the SO2 tower before entering the electrolytic system. Under these conditions, only 0.56 lb. of sulphur was required per pound of ferric iron reduced.
Although the small tower was not of sufficient capacity to make control easy, by Aug. 22 the ferric iron was on the down grade—1.07 per c.ent. as against 1.22 on July 22.

From this time on, as long as lead anodes were in use, the ferric iron did not get out of control, and on Nov. 5 the solution showed only 0.24 per cent, of this troublesome substance. The results with graphite anodes will be considered later.

Fouling of Solution

In the 40-ton plant the entire solution began to foul badly after about 100 charges had been leached. The cooler weather was beginning at this time and a bulky double sulphate of iron and copper crystallized out through the ore in the leaching tanks and plugged all the pipes and launders. To remedy this, a portion of nearly neutral solution was drawn off continuously from the leaching system and passed over scrap iron, making cement-copper. It was found that treating 1 per cent, of the total bulk of solution each day would maintain the fouling at a constant low point where it could cause no trouble. Columns V and VI of Table 4 show the effect of this “bleeding” of solution.

The cement-copper so produced was returned to a wooden tumbler and electrolyte on the way to the cells passed through it. Ferric iron is reduced to ferrous in this way and copper passes back into solution.

Under these conditions only electrolytic copper is produced and at the same time the control of the ferric iron is aided considerably. In the 40-ton plant it was apparently necessary to send about 12 per cent, of the total copper through this side cycle and back into the main one, to control completely the fouling materials introduced from the ore during leaching.

Grade, of Copper Produced

Cornelia copper produced direct from the ore is the equal of any electrolytic copper produced in this country. Fortunately the orebody contains only minute amounts of either arsenic, antimony or bismuth. The analysis and the physical tests on Cornelia cathodes will be found in Table 5.

Size of Material Percolation

The work at Ajo was all done on the product from a Symons fine crusher, without any screening (except for tests on sized material). It was nominally a 4-mesh product but carried an average of about 30 per cent, of oversize. Extraction on this is satisfactory and while it could be increased by finer crushing, a large percentage of the copper minerals lies in fracture planes, and no very great profit would result.

This product is especially well suited to upward percolation in a bed 12 ft. deep or more. It does not segregate nor channel badly, drains well and is easy to wash.

Circulation in the Leaching System

In the 40-ton plant, each of the eight active tanks had a closed circulation, upward through the ore, of about 80 gal. per minute. This rate is probably much higher than necessary, but for the experimental work it was made ample and retained as a fixed factor. Beside this closed circuit, which runs continuously, a portion of the solution from each tank is continuously advanced to the next tank in the series.

Fresh electrolyte from the cells, low in copper and of maximum acid content, leaves the electrolytic system and passes into the leaching tank containing the oldest ore—that which has been leaching for 7 days. Practically neutral solution, high in copper, leaves the leaching system at the tank containing new ore. Between these two points, and rigidly linked to them, is the advance from tank to tank.

It is of great importance to have a practically neutral solution leaving the last leaching tank and starting back toward the electrolytic system.

At the same time, we want as high an extraction as possible, which means as many tanks as possible with a fairly high acid content. To attain both these objects at the same time the advance from tank to tank is carefully regulated so that about six tanks of the eight active ones are acid practically all the time. The two others are neutral and the point of neutrality of the system just balances, varying possibly one tank in either direction during 24 hr., but returning to its original position.

Circulation in the Electrolytic System

The electrolytic cells, together with tanks and sumps for storage, form a second closed circuit from which solution is bypassed continuously into the leaching system at the same rate as the advance from cell to cell. To produce good, coherent cathodes, the circulation through the cells must be much greater in volume than the advance from the leaching system, as the latter is rigidly fixed by the necessity for maintaining neutral leaching tanks. While the advance in the 40-ton plant was about 12 gal.

per minute, it was advantageous to circulate in the tank house at rates up to 150 gal. per minute.

This circulation gives satisfactory results with lead anodes, but when graphite is used, rapid circulation alone is not enough to give good depolarization at the anode. Violent agitation in the electrolytic tanks is necessary to attain the lowest voltage when graphite anodes are in use, and this was accomplished by blowing air through small holes in lead pipes laid along the bottom of the cell.

Lead vs. Graphite Anodes

In making final decision on the process to be used, there were many mechanical questions to be settled. Crushing, method of percolation, circulation, pumps, structural materials, the handling of ore and tailings and many other points came up for decision. Besides these, there were some questions which bear more directly on fundamental principles. One of these was whether we should use lead or graphite as material for anodes.

Lead was in service for over a year and its behavior was well understood. But graphite promised much better results as far as power efficiency is concerned, and also as a means of regenerating a much larger part of the necessary acid for leaching.

The 1-ton plant gave most encouraging results with graphite over a period of continuous operation of nearly 3 months. With a current density of 8.5 amp. per square foot, over 2 lb. of copper per kilowatt-hour was produced, and during part of this time conditions were not good. The ferric iron was not under complete control. When all conditions were right, we made 2.25 to 2.40 lb. copper per kilowatt-hour; on some days as high as 2.75 lb. This was during the warm weather at Ajo, when electrolyte temperatures held well above 105° F. The 40-ton plant was using the large tower for reduction and the 1-ton plant used the small one. The SO2 used in controlling ferric iron was bypassed from the main supply.

The 1-ton electrolytic plant was out of doors and while the odor of SO2 was strong in the immediate neighborhood of the cell, it was not bad enough to prevent working over it.

By the time the graphite anodes were installed in the 40-ton plant, the average electrolyte temperature was only about 75° F. We could now determine accurately the efficiency of absorption and reduction by SO2 (which we could not do in the 1-ton plant) and we could also examine carefully the circulation conditions necessary to control ferric iron.

The problem now began to look harder. As measured by the production of ferric iron from ferrous in the cell, the anodic efficiency of graphite, with rapid air agitation, is well over 100 per cent. With lead it is 30 to 35 per cent. We had, therefore, to combat about three times as much ferric iron per pound of copper with graphite as with lead.

Under the existing temperature conditions and with the rapid circulation that was necessary in the electrolytic circuit, the reduction efficiency of the SO2 was only about 20 per cent. We could not control the ferric iron without either allowing the electrolyte to stand for a good many hours between the SO2 towers and the cells, or else heating it to about 150° F. during its passage from towers to cells.

The absorption of SO2 in the towers was perfectly satisfactory. But unless it was used to better advantage in the reduction of the iron in the solution, it was impossible to live in the tank house without either very careful ventilation or a diver’s outfit. With the violent air agitation that was necessary for high power efficiency, most of the SO2 which had been absorbed in the towers was blown out in the tank house.

These difficulties are by no means insuperable. Of more fundamental nature is the low current efficiency obtained with graphite and its extreme sensitiveness toward ferric iron. When ferric iron is practically completely removed by careful reduction with SO2, graphite can be made to give high current efficiencies. But the very slightest rise in ferric-iron concentration results in an immediate and considerable lowering in ampere efficiency. Even in our good runs with graphite in the 1-ton plant, the current efficiency was not much above 70 per cent. With lead anodes, and without any particular attention to the ferric iron, which was usually 0.3 or 0.4 per cent., we often obtained current efficiencies of 90 per cent, or better over considerable periods.

It may be of interest to state some of the points that enter into a decision between the two kinds of anodes:

1. The first cost of an installation is about the same.
2. The life of lead is known with some accuracy. The life of graphite in this service is not known.
3. The salvage value of lead is high. With graphite it is practically nothing.
4. The probable power saving to be expected (graphite over lead) if graphite were successfully operated, is about 0.6 kw.-hr. per pound of copper.
5. More than twice as much acid is regenerated with graphite as with lead.
6. Circulation must be more rapid with graphite.
7. Agitation during electrolysis is absolutely necessary with graphite.
8. The solution must stand for some time (probably 16 to 24 hr.) or be heated to 150° F. or higher, to obtain even reasonable SO2 efficiency with graphite. This is not necessary with lead (see note under 10).
9. Much larger towers are needed when graphite is used.
10. The volume to be pumped to the top of the towers is larger with graphite.
    (These last two statements contain the fact that acid electrolyte must be sent through the towers. The iron content of the solution in the cells must not rise above 0.2 per cent., if good efficiency is to be reached with graphite. With lead anodes it may go to 0.4 or 0.5 per cent, without much of any effect on power consumption. With lead it is sufficient to send the neutral advance to the towers. With graphite not only this volume, but also a much larger volume from the tank-house circulation which must be kept low in ferric iron must be sent.)
11. The tank house for graphite must be carefully ventilated.
12. The tank house for graphite must be larger (about as 9 to 7) on account of the lower current efficiency.
13. Operation with graphite requires the closest attention to every detail of reduction, circulation and agitation. Operation with lead is practically fool-proof.

This last point is probably the most important of all. There are enough so-called “minor” troubles about a leaching plant.

Control of Ferric Iron in Acid and Neutral Solution

During the periods when acid electrolyte was being sent to the tower for the reduction of ferric iron by sulphur dioxide, the sulphur efficiency was low. It proved to be about 20 to 25 per cent. Two causes at least combine to produce this result: First, the low solubility of SO2 in acid solution, and second, the slowness of the reduction reaction in acid solution. Neutral solution was sent to the towers in later runs and on this the sulphur efficiency reached 75 to 80 per cent. That is to say, of the actual weight of sulphur burned to SO2, 75 per cent, is utilized in the reduction of ferric iron to ferrous.

Washing the Ore

Each charge was given three washes. The first was the second wash of the preceding charge; the second, the third wash of the preceding charge; the third, fresh water. The wash which had been used three times was run into the electrolytic system. One complete fresh wash water was just sufficient to balance the loss from the system in the tailings and by evaporation.

The summarized wash-water data are given in Table 6 for the solution which was started in the 40-ton plant May 22, 1915. The most significant thing in this table is the effect of fouling as shown in the second column. While the solution was fresh, washing in the prescribed manner, the extraction was 81.1 per cent, and soluble copper left in the tailings was 0.045 per cent. When the solution became foul, even though a little more fresh water was used, extraction dropped to 73.5 per cent, and the soluble copper in the tailings was 0.103 per cent. Later, when fouling was being reduced by “bleeding” and additional wash water could be used in the same cycle, the extraction rose to 83 per cent, and the soluble copper in the tails dropped to 0.033 per cent.

It is of course possible to wash tailings with extra water, passing this over scrap iron and making cement-copper, and this should naturally be done until the expense of operation balances the value of the recovered copper. The experiments at Cornelia were carried no further than is shown in this table.

All the experimental work in the 40-ton plant was carried out during continuous operation, so that the result should correspond as closely as possible with commercial conditions. The regular tonnage was crushed each day and charged into the leaching tank. Tailings were removed, and circulation, advance, washing, electrolysis and reduction carried on without a break. Any change in operation was carried on for at least a week, and usually for 10 days. If the change promised to give improved results, it became a fixed part of the regular operating routine and the next point came up for study.

Using this method, when we were through, we had a definite set of conditions under which known results could be produced for any length of time. There were no odds and ends to go back and pick up. No “minor” problems, which might turn out on investigation to be fundamental, remained to be solved. Mr. Greenway saw that this plan was strictly carried out.

During the experimental work, something over 180,000 lb. of perfect cathodes were produced, besides a good deal of cement-copper.

Summary

So far as the Ajo tests can show what will happen in a 5,000-ton plant, the following may be expected:

Extraction on 1.4 per cent, ore using closed wash cycle, 82 per cent.
Extraction on 1.4 per cent, ore using extra wash water, 83 per cent.
(The extraction figure used in making estimates has always been 80 per cent.).

Acid consumption (100 per cent. H2SO4), 3 lb. per pound of copper. “Bleeding” to prevent fouling of solution.

Power efficiency in electrolysis with lead anodes, 1 kw.-hr. per pound of cathode copper.

Sulphur consumption, 0.5 lb. per pound of cathode copper.

Neutral electrolyte only will need to pass through the reduction towers.

About 1 per cent, per day of the total solution volume must be sent over iron to control fouling.

About 12 per cent, of the total copper produced will pass through the cement-copper side cycle.

The circulation used in the test-plant leaching tanks (2 gal. per minute per ton of ore) is undoubtedly much higher than necessary. It is probable that a small part of this volume would give equally good extraction results.

Discussion

The Chairman (H. W. Morse).—Gentlemen, for the first time in the history of the American Institute of Mining Engineers, we have a full session on the subject of leaching—especially on the leaching of copper ores. This branch of metallurgy is rather in its infancy. We have had one big plant running for a year or more, and pretty successfully; and another one which will be started in February or March, the New Cornelia. We have another one at Garfield for the Utah company, which is just designed and construction begun; and from several other points come reports of interesting test work. The first paper on the program this evening is one of some of the test work on the New Cornelia ore by myself and Mr. H. A. Tobelmann.

Lawrence Addicks, New York, N. Y. (written discussion).—The Ajo proposition has had, as I understand it, two great advantages: it has not been obliged to measure leaching against flotation or some other metallurgical process, and it deals with a copper content high enough to pay for straight sulphuric acid-iron cementation work if forced back upon such a plan.

These advantages have made it possible to prosecute a consistent policy of development along straight hydrometallurgical lines.

The history of these experiments shows a series of investigations into more or less complicated schemes with a gradual return toward simplicity, the present plan involving utilization of SO2 in regular acid making practice, precipitation of excess impurities upon the ore itself, final control by cementation and recovery of the copper by the use of lead anodes along more or less copper-refinery lines.

This is about as simple a scheme as could be devised in true cyclical form, and while I think no one will dispute the wisdom of adopting just such a scheme as the first step in large-scale operation, I venture to predict that as experience is gained in actual practice the tendency will be to return to greater complication, as one item after another can be examined separately and tested for cost. The great difficulty in launching a leaching process lies in the multiplicity of interrelated problems which must at first be dealt with jointly. Also in the fact that no two ores seem to present identical problems for solution.

I think this feeling I have outlined is exemplified by the discussion of the reasons for choosing lead anodes as given in the paper and I want to add a few words to this discussion.

The crux of the whole electrolytic problem lies in the mechanism of the oxidation and reduction of iron sulphate. A carbon anode must be thoroughly depolarized or the liberated oxygen will attack the anode itself and the graphite will disintegrate leaving a valueless sludge. In the case of a lead anode, lead peroxide and lead sulphate are formed but the disintegration is very slow as these substances form more or less adherent and conducting coatings. This means that with carbon anodes we impose at once the necessity of dealing with the entire equivalent of ferric sulphate if ferrous sulphate be the depolarizer. With lead anodes the issue can be largely but not wholly dodged through the escape of elemental oxygen. This in turn leads to the necessity of thorough circulation in the case of carbon anodes as against indifferent circulation in the case of lead.

Thorough circulation, however, results in intimate contact between the cathode and the ferric sulphate in the electrolyte with the natural and undesired reaction between these two substances.

Therefore, the price of our increased recovery of copper per kilowatt-hour is the obligation to keep liquors running more than some 0.2 per cent., and preferably much less, in ferric sulphate, away from intimate contact with the cathode, which means either a diaphragm or an efficient utilization of SO2 or other reducing agent.

Everyone who has advocated the first remedy has come to grief, except possibly Hybinette in his two-level system in copper nickel refining, and everyone who has wrestled with the second problem has found that there are intermediate and as yet little understood reactions to be dealt with. Nevertheless, I believe that as continued operation gives experience and opportunity for experiment, a way out of these difficulties will be found and the carbon anode yet succeed in the practical electrolysis of sulphate solutions.

G. D. Van Arsdale, New York, N. Y.—Both those responsible for the tests of which the results are given in this paper and the Institute are to be congratulated; the first for the successful results finally obtained on a difficult and new problem, and the profession on the liberal policy which made possible the publication of this very valuable summary of practical leaching data.

The thoroughness and care with which this work has been done may be judged from the time, nearly 4 years, during which these experiments have been carried on.

It is especially gratifying to the writer that some of the main points developed during the series of experiments on leaching by Phelps, Dodge & Co. at Douglas have been entirely borne out by this work; the first of these being that ferric iron can be controlled by the use of sulphur dioxide, the second that solutions that had previously been considered entirely too “foul” as an electrolyte for copper deposition can safely be used, and the third that by the use of depolarization the anode problem is entirely solved, either by lead or graphite, both of which had previously been considered as impossible for electrolysis of sulphate solutions of copper.

The paper is particularly valuable in giving a discussion of the conclusions leading to the adoption of the final process by the New Cornelia company and as bearing on their choice of lead instead of graphite as an anode material. My reason for discussing, and to some extent differing from, these conclusions is that I feel that from the results of Phelps, Dodge & Co.’s work it is necessary to emphasize the fact that, while the choice of lead in this case may be undoubtedly justified, yet there are conditions under which graphite should be selected and these conclusions therefore should not be taken as general conclusions against its use under other sets of conditions.

Graphite, which was first used under the writer’s direction on a considerable experimental scale as an anode material for this kind of work, before this time had been considered as absolutely unsuitable for the electrolysis of sulphate solutions of copper. I have a letter from the manufacturers in which this statement is made, and the same thing is stated in Greenawalt’s book on copper leaching. There is no doubt that its use has certain practical disadvantages, the principal one of these being that a tank-room ventilation system will probably be needed to enable it to be used. While this means a radically different tank-room design, it need not mean anything at all impracticable or leading to increased tank-room costs beyond the negligible amount of power for ventilation. On the other hand, its use has very considerable points in its favor which in my opinion outweigh those against it, generally speaking.

Regarding its durability, I believe it is entirely safe to say, from the results of careful tests made by Phelps, Dodge & Co., that under proper and easily controlled conditions its depreciation costs, or in other words the amount of disintegration, per pound of copper produced will be negligible.

The ampere efficiency of graphite is, as stated, badly reduced by a ferric iron content of electrolyte, with which still fair efficiencies may be had with lead, this meaning as stated the maintaining of a limit of about 0.2 per cent, ferric iron for graphite as against a limit of about 0. 4 per cent, with lead. It should be carefully noted, however, that as regarding the amount of power per pound of copper produced a reduced ampere efficiency at a voltage of about 1 volt does not mean nearly as much as the same reduction with a voltage of over 2 volts. In other words, one of the very considerable practical advantages of graphite is the much lower voltage at equal current densities with lead. Regarding the ampere efficiency obtainable with graphite, it is possible that a casual reading of the paper under discussion might lead to the conclusion that a low ampere efficiency was characteristic of graphite. That this is not true was proved conclusively by Phelps, Dodge & Co. in a series of tests last summer which demonstrated that it is commercially practical to keep the ferric iron under 0.2 per cent, and under these conditions to obtain fully as good or better efficiencies than those cited in the present paper, with yields over a considerable period of 3 lb. of copper per kilo-watt-hour at a current density of about 12 amp. per square foot, which it should be noted is 76 per cent, in excess of the average current density as given in the table on page 837 of the paper under discussion. With the lower current densities as given in this paper, our pounds of copper per kilowatt-hour would have been still higher.

It is, of course, not reasonable to expect that a high SO2 reduction efficiency will be obtained if the SO2 is blown out of the solution before it has had time to act. In the worst case, this factor of the time needed for the reduction of ferric iron by SO2 means a large, though not at all prohibitive, storage capacity, and it may be stated that one of the results of our work at Douglas last summer was to demonstrate that the ferric iron can be kept within our 0.2 per cent, limit by a storage of a portion only of the main solution.

It is true that the anodic efficiency of graphite is as stated, on account of air agitation, over 100 per cent., that of lead being 30 to 35 per cent. Expressed in another way, however, this means, with anything like a reasonable SO2 reduction efficiency, which can be had as we have shown, that for practical purposes the amount of acid regenerated by lead is much less than with graphite. Now if one has an acid plant already built or ordered or intends to erect one, the comparatively small amount of acid produced by the use of lead anodes is not so important, but where this is not the case this considerable extra amount of acid produced by the use of graphite is a considerable practical advantage in its favor. Where, as in this case, an ore does not require more than 3 lb. of acid for leaching per pound of copper produced, the amount of acid regenerated by the use of the graphite system will be ample for leaching purposes. A comparison, therefore, between the two systems should include the cost of an acid plant against lead as regards installation costs, and, as regards operating costs, the extra cost of acid needed together with freight if brought from a distance. It seems to me, therefore, that the statement in point 1 of the summary on page 844 should include this factor.

Without intending criticism of the decision made in this particular case, I believe, nevertheless, that the following comparison as to the relative merits of lead and graphite will be true and more generally applicable than those given.

1. Installation costs for the two systems will be about the same, except that, where an ore requires more than 1½ lb. of acid per pound of copper produced, the additional costs of an acid plant plus railroad equipment for carrying it when produced at a distance from the leaching plant, plus storage tanks, piping, etc., at the leaching plant, must be charged against lead.

2. Careful determinations extending over a considerable period have shown that the disintegration of graphite and its consequent depreciation per pound of copper produced is negligible. The prevention of graphite anode disintegration means simply maintaining the conditions necessary for proper depolarization. When this is done graphite will not disintegrate appreciably, but if these conditions are not maintained it will go rapidly. Exactly the same thing is true for lead, but, since the anodic efficiency of lead is only one-third that of graphite, the disintegration of lead anodes under the same service may be expected to be at least as rapid as graphite.

3. Peroxidized lead detached from an anode will be lost, and the scrap value of partly peroxidized and badly corroded lead anodes, when high freight rates are considered, will not be very high. Furthermore, when the slow rate of disintegration of graphite anodes finally reduces their thickness very appreciably, they can be reassembled and the scrap loss reduced thereby to a small amount.

4. The power required for precipitation with lead is at least three times that for graphite under the same conditions.

5. More than twice as much acid is regenerated with graphite as with lead. This, as stated above, means not only an extra installation but also an extra operating cost for this additional acid needed, when lead is used.

6 and 7. Since the anodic efficiency of graphite is very high, higher circulation and some form of agitation in the cells are required. In this connection it may be stated that, although air agitation has been used in the work at Douglas with good results in some respects, I have never been in favor of this method of agitation as compared with others, unless more than 3 lb. of acid per pound of copper are needed. The excess anode efficiency over 100 per cent, is of course mainly due to the oxidation caused by the air, and it is clearly illogical to use an oxidizing agent for mechanical purposes where reducing conditions are wanted.

8. The extra storage capacity needed for the time of reduction by SO2 as shown above need not be more than a fraction of the total solution bulk. Furthermore, if a good SO2 reduction is obtained, the tank-room ventilation difficulties are correspondingly reduced.

9. and 10. Since the anodic efficiency of graphite is three times that of lead, it of course necessarily follows that larger tower capacity and greater volume of solution sent to these are required. However, the saving in installation by not needing an acid plant will pay for this and the extra storage capacity needed under the preceding paragraph at least several times over.

11. The point of the necessary tank-room ventilation has already been spoken of.

12. I must disagree absolutely, so far as my experience goes, with this conclusion as to tank-house capacity being necessarily larger with graphite than with lead. While the cost of power is an important factor in this connection, we have not considered for our requirements that the low-current densities, averaging 6.8 amp. given in the paper, are economical under our conditions, and the figures we have obtained with current densities around 12 amp. per square foot, in which average ampere efficiencies at this higher density are fully equal to or better than those given in the paper and obtained over considerable periods appear to throw this comparison decidedly in favor of graphite.

13. It is a little difficult to see just why operation with graphite anodes will be any less “fool-proof” than with lead. Granted that the conditions for keeping the ferric iron below our limit are known, I cannot see any reason whatever for requiring any more care or skill for maintaining this limit than for maintaining an equally vital one differing from this by only 0.2 per cent.

My opinion, not expressed as a criticism of the decision made in this particular case, but as a general one, is that the advantages and undoubted lower operating costs of graphite decidedly outweigh those with lead. In other words, it would only be considered advisable under very exceptional conditions to make the very large additional investment required for an acid plant over and above the acknowledged nearly equal costs of the two systems, and still with this extra investment to obtain a higher operating cost.

The Chairman.—I would like to say that Mr. Van Arsdale should be thanked for a good many discussions and arguments, and that the conversations and arguments with him were of very great advantage to us in this work at the New Cornelia; and in a good deal of what he says I would agree. It is highly probable that we will be able some day to work out the use of graphite to make a saving in power cost, and to generally improve the process. As far as we could see from the test at the New Cornelia, this was going to be a difficult undertaking, so we decided to put it off for a few years. I think we will later come back to large-scale tests on the use of graphite. Certainly the gain to be expected when this has been worked out is a great one. I, personally, was sorry to stop the research work on the graphite.

S. J. Jennings, New York, N. Y.—I would like to ask for some information. I see that this paper has considered merely two kinds of insoluble anodes, lead and graphite. I understand a considerable amount of work has been done with other insoluble anodes and I would like to ask if tests have been made at New Cornelia with any other insoluble anode such as fused magnetite.

The Chairman.—We have made none at Cornelia, except with lead and graphite, and a mixture of the two—that is, coke cast in lead.

F. S. Schimerka, Clifton, Ariz.—In connection with the paper presented, I wish to ask the question why the process of precipitating the ferric iron in a neutral liquor on fresh ore was not finally employed and preference given to the introduction of sulphur dioxide to effect the elimination of trivalent iron?

In regard to the use of lead anode sheets, I would like to know whether lead-covered iron grates would be applicable, and I ask this question because I remember that I have seen such anodes employed in liberator tank practice but have not been able to collect data concerning their efficiency compared with that of solid sheets used for the same purpose.

The Chairman.—That was the basis for starting the large plant. In the 7-ton plant and in the 40-ton plant, at first, the ferric iron was precipitated, but later for an unknown reason we got no such result, the ferric iron going up and the ampere efficiency going down.

C. G. Grabill, Matehuala, Mex.—I would like to ask if it was possible that the difficulties arose from the depth of ore in the two tanks. In the first one you have a small volume of solution; and in the second, you have a very much larger body of solution. In the first condition, with the small volume, you have much more air present and oxidation taking place. If the charge is increased—rather, if the percentage of the solution is increased—that condition does not obtain in the latter part of the action, I would like to ask if you have done any work along that line—to investigate that point?

The Chairman.—That is one question we left to find out in the 5,000-ton plant. Mr. Flynn, have you anything to say?

F. N. Flynn, Clifton, Ariz.—The leaching that we did at Clifton, for something over 20 years, was the simple process of an acid leach, with iron precipitation, and the wasting of all liquors. The modern leaching which you are about to attempt is something which you have developed and proven to your entire satisfaction and until such time as it is working out, practically, I don’t feel that I would have any criticism to make. I don’t feel competent to criticize. It does seem to me that you are working in the right direction. The only question that comes to my mind is the percentage of liquors you will have to throw away. That, I understand, you have worked out to your entire satisfaction.

The Chairman.—I think that 6 months’ operation of the big plant will make us all feel more confident than we do now.

Stuart Croasdale, Denver, Colo, (communication to the Secretary).—I think the New Cornelia ores at Ajo may be considered as typical of all copper ores in the United States that are amenable to acid leaching in a raw condition, since all known ores of this kind contain a certain amount of soluble iron oxides and alumina which will be a contending factor in electrolytic precipitation of the copper.

It is interesting to observe that the results obtained by the authors have verified in almost every instance the predictions made before they started, yet their work has been so efficiently and thoroughly done that it forms a most valuable contribution to the hydrometallurgy of copper, and particularly to the section on electrolytic precipitation. They are to be congratulated on having such an opportunity to definitely advance our knowledge so much in this direction.

I regret that one thing more was not tried, at least on a laboratory scale, and that is, roasting the ore before leaching. In the Ajo ores the iron exists in all forms of oxidation from chalcopyrite to ferric oxide. The processes of nature have been so gentle that much of this iron is as easily soluble as freshly precipitated hydroxide. Aluminous compounds in the rock also exist in all stages of kaolinization and are likewise soluble in dilute acids. A slight roast would oxidize all of the iron, including the residual pyrite, to ferric oxide and would dehydrate the aluminous compounds and salts to the oxide. Since ignited ferric and aluminum oxides are not readily soluble in dilute acid, it would seem that the lixivium obtained from roasted ore would be practically free from deleterious salts, which would obviate the necessity of depolarizing and “bleeding.” Assuming this to be true and the solubility of the copper oxide not diminished, the cost of roasting would be compensated by the recovery of the copper now held in the tailings as sulphide, which is likely to be in ever increasing quantity, and as soluble copper retained by the argillaceous material; also by the expense of roasting pyrite to produce sulphur dioxide, and by the expense of precipitating part of the copper on scrap iron at each cycle of the lixiviant.