Electrochemistry of Flotation

If one turns to ‘Elementary Lessons in Electricity and Magnetism,’ by Silvanus Thompson and studies the fundamental principles of frictional electricity, as given in Chapter 1 of Electrochemistry of Flotation, a clearer idea of the causes of ‘flotation’ may be obtained. After seeing a few experiments, such as were performed at the Case School of Applied Science early in the year, it is not a difficult matter to believe that most of the phenomena are electrical in nature. For instance, if powdered galena ore, with a limestone gangue, be dropped into pure water, most of the powder will immediately sink to the bottom. As the air enclosed by the particles is expelled gradually, one sees the formation of ‘armored’ bubbles, some of which may last for days. Here is flotation without oil or acid. If nitric acid is added, the gas bubbles, formed by the action of the acid on the gangue, will carry up particles of galena, some reaching the surface and bursting, while others too heavily loaded with galena particles will hover just below the surface. These will form clusters, resembling bunches of grapes, and when enough gas bubbles join the clusters, they start upward toward the surface, but generally before reaching there they are overloaded by particles falling from the bubbles that are bursting at the surface. The bubbles with their loads often resemble balloons, with the galena hanging on to the bottoms, as do the baskets of actual balloons. Some of the bubbles will be completely ‘armored’ while others will be nearly free from galena. Another experiment that may be successfully used in the laboratory for the flotation of the difficult sulphides, such as old rusty pyrite concentrate, like sweepings from floors of old mills, is as follows: Mix the ore with bleaching powder, some carbonate (say, sodium carbonate), and water. Put the mixture into a glass beaker and add concentrated nitric until red nitrous fumes are given off. Chlorine also will be evolved. The bubbles of gas are so highly charged electrically that pyrite from the Mother Lode between 10 and 20-mesh size was floated, making a complete separation from the quartz gangue. In this experiment the nitrous oxide was the active agent, for if the same experiment is conducted with sulphuric acid, no such separation takes place. Assayers are familiar with a similar phenomenon when treating blister copper with concentrated nitric acid and heating. Nitrous oxides are formed and the metallic copper is floated, a froth of copper being the result. Carbon dioxide does not seem to be as active as the nitrous oxides or chlorine bubbles.

ELECTRIFICATION OF THE BUBBLE

Two different substances, whether gaseous, liquid, or solid, when brought intimately into contact and moved one over the other, always produce electrification. Difference of temperature of two similar substances in frictional contact will cause electrification, the warmer usually being negatively charged. Something certainly happens when the surfaces of two different substances are brought into intimate contact, for the result is that when they are drawn apart, they are oppositely charged. The nature of the charge depends on the substances. Fur rubbed on glass electrifies the glass negatively; while if glass is rubbed with celluloid it will become charged positively.

A blow struck by one substance on another produces opposite electrical states on the two surfaces. Again, the evaporation of liquids is accompanied by electrification, liquid and vapor assuming opposite charges, though this is only apparent when the surface is in agitation. A few drops of copper sulphate thrown on a hot platinum plate produces violent electrification, as the copper sulphate evaporates. Electrical charges are set up by various other means, such as vibration, disruption of material, crystallization, combustion, pressure, and chemical reactions.

It would seem easier therefore to electrify a bubble than to keep it from being electrified. I assume that the bubbles are electrified, whether by means of air being forced through canvas, by beating air into water with blades, or by other means. The next step is to consider the properties of an electrified sphere. These may be illustrated by suspending two light spheres of conducting materials near each other by means of silk threads. Upon charging the spheres with like electric charges, they will repel each other, but if a conductor is brought toward them, both are attracted to the conductor. Of course, if the spheres touch the conductor and the conductor is grounded, then the spheres lose their charges. If the conductor is insulated from the ground, then upon contact with the spheres, the conductor receives a similar charge and the spheres will be repelled. Suppose that the spheres or conductor are covered with an insulating film. Then the spheres and conductor would remain as close together as the films would permit. So air bubbles that are electrified will attract conductors near them that are free to move.

Air, being a poor conductor of electricity, the bubbles as a whole do not discharge immediately upon contact with a conductor. The only part of the surface discharged is that in immediate contact with the conductor, and this discharged film of air acts as a dielectric and non-conductor to the rest of the bubble, which remains charged.

The amount of electrification of the bubble will depend on various conditions, such, for example, as the amount of friction produced by the blades of a Minerals Separation machine. Increase the speed and the electrification is greater and the attraction for conductors will increase, reducing the proportion of conductors in the tailing. Referring to D. G. Campbell’s article in the Mining and Engineering World of January 17, 1914, the speed of agitation and the percentage of extraction is given as follows:

The extraction seems to vary directly as the square root of the increase in speed. But it will be observed that with the increased extraction, the percentage of sulphide in the concentrate decreases, due to the attraction of the small particles of mixed gangue and sulphide. If the bubbles are highly charged, the concentrate will not be as clean in a particular case, as if they were less charged.

Vapors and gases may be highly electrified. The Armstrong hydro-electric machine, devised by Lord Armstrong, gave sparks of 5 to 6 feet. The friction of a jet of steam through a wooden nozzle generates the charge on the particles of condensed water.

RELATIVE CONDUCTIVITY

From the above, it appears that to float a mineral, it must be a conductor. The following table of relative conductivities is taken from Landolt-Bornstein ‘ Physikalisch- Chemische Tabellen,’ 1912, fourth edition:

From this table, it would seem that the metals and sulphides that may be recovered by the flotation method are all conductors. The chalcopyrite figure seems low, but the flotation properties of sulphide minerals vary, and the variation in the conductivity of the different minerals may have something to do with this. The better the conductivity of the valuable mineral, the easier is it floated, other factors remaining the same.

THE INSULATING FILM

The next important question in the problem is the action of the oils, resins, or other agents now used in flotation. Working from the electrical standpoint, it is necessary to prevent the charge of the bubble from being dissipated and thus breaking down the froth, before it has done its duty. Oils and other substances have a tendency to coat the metals and minerals that are recovered by flotation, and if the air bubble is completely surrounded by these particles, an envelope of oil or other dielectric will insulate the bubble and prevent the dissipation of the charge. Without a dielectric film about the bubble no permanent froth would be formed. It is, therefore, necessary to add some material of great dielectric strength that has the tendency to coat the valuable mineral.

The words ‘dielectrics’ and ‘non-conductors’ or ‘insulators’ should not be confused. A ‘dielectric’ is a substance that is not only a non-conductor, but is also one that takes part in the propagation of the electric inductive forces. All dielectrics are ‘insulators,’ but equally good insulators are not necessarily equally good dielectrics. Air and glass are far better insulators than ebonite or paraffine, but the inductive influence acts more strongly across a slab of glass than across a slab of ebonite or paraffine of equal thickness, and better still across these than across a layer of air of the same thickness.

It may, therefore, be possible to use a frothing agent, as is well known, that is not an oil at all. I have done this and formed froth that has lasted for weeks. For instance, if in the experiment mentioned in the first part of this article, with galena ore, a little alcohol is first mixed with the galena, before the water and acid is added, then a heavy mass of bubbles and galena particles will be formed, too heavy to rise to the surface.

As the influence of the charge acts inversely as the thickness of the film, it is imperative that some dielectric be used that will create a very thin film about the valuable mineral. The dielectric must also be of such a character as to aid the formation of a great quantity of small bubbles in the liquid. It is difficult to create and maintain small bubbles in pure water. It is here that surface tension phenomena probably play a part in flotation.

ACIDITY OF THE PULP

In Mr. Campbell’s article he gives the following results of acid variations:

Other tests also show that the extraction decreases as the acidity increases, but the amount of gangue in the concentrate decreases much more rapidly. With an acidified pulp, a cleaner concentrate is obtained. Also better results are obtained if the acid is added before the oil to the agitation-tank.

As to the action of acids and alkaline substances in the pulp, little seems to be known, but according to the electrical theory, the addition of these substances causes the conductivity of the pulp to increase greatly. It is a possibility that if the acid is not added before the oil, the gangue, oil, and conductors are all electrically charged by reason of the friction. The conductors would be positively charged, while the other substances and the air bubbles would be negatively charged. If the pulp is a poor conductor, as it would be if water is not acidified or otherwise made a conductor (pure water being a very poor conductor), the charges on the gangue materials would remain for some time and the conductor (sulphides, etc.) would attract the gangue as well as the bubbles and oil, thus causing gangue to be taken up with the bubbles. By the addition of acid, the charges on the surface of the solids are discharged to the ground, and the bubbles and the oil, which will not be instantly discharged as are the solids, will attract the conductors.

CONCLUSIONS

It might be stated here that the electrical theory was taught last year, as possibly explaining flotation phenomena, to the class in ore-dressing, at the Case School of Applied Science.

The above mentioned method of floating conductors may be used for the rapid determination of certain ingredients in ores that are amenable to the flotation process. It requires only a beaker and a few chemicals, no flotation machines being needed. For the rapid approximate determination of insoluble in a smelting ore, the method will give a fair result within a few minutes. If the conductor is readily acted upon by nitric acid, the results may not be satisfactory, but by addition of oleic acid the dissolving action of the acid is reduced.

The following summary of the requirements for ‘flotation,’ considered from the electrical standpoint, may be of practical use:

1. Ores containing valuable minerals or metals that are good conductors are the only ones that are suitable for flotation.
2. To buoy these conductors, it is necessary to supply enough electrified bubbles from below to float particles of the conductors that are attracted; hence the smaller the bubble, the better the result, the amount of gas being the same.
3. Some dielectric fluid is necessary to cover the conductor or the bubble, to prevent the dissipation of the electric charge. The thinner the film of dielectric and the greater its dielectric strength, the greater the effective attractive force and the more permanent will be the froth.
4. Some material must be added to the water to increase its conductivity, to obtain a clean concentrate; acids in small quantity are now used.

It seems to me, however, that the fundamental principles of flotation can best be studied with larger particles, thus avoiding the interesting, but also little undestood ‘colloid’ chemistry. In gravity separation by rising currents of water, Rittinger’s formula V = c√D(S-L) holds true for particles above a certain size, namely, about 0.2 mm. for quartz and 0.13 mm. in case of galena. Below these sizes, colloidal and other little-known phenomena become of importance and complicate the investigation. So it is with flotation.

In the laboratory, it is possible to use larger particles. The following experiments were conducted in the Case School of Applied Science on material sized through 20 and 30-mesh screens. The phenomena connected with preferential flotation furnish new evidence to strengthen the electrical theory.

The simplest experiment, demonstrating preferential flotation, may be performed as follows: Upon a 4-inch watch-glass, place a little galena, blende, and quartz, of 20 to 30-mesh size. Add dilute nitric acid and place the glass under a miscroscope. The acid attacks the galena, forming bubbles of H2S gas that adhere to the galena. The particles of galena are electrified also, as can be seen by the actions of the particles. The blende and quartz are not attacked. If the ore had been finely pulverized and dilute nitric acid added, the bubbles of H2S would have been sufficient to float the galena, leaving the blende and quartz at the bottom. However, with fine particles, some blende and quartz would have been entrapped, brought to the surface, and held there by surface tension. The bubbles are not sufficient to float the coarse galena, but by a vanning motion of the glass, the galena will collect, being brought and held together by the H2S bubbles, forming a mat, which is lighter than quartz or blende and can, therefore, be panned off, leaving the blende and quartz. This experiment seems to show that the H2S is charged oppositely to the galena.

If more concentrated nitric acid had been added to the ore, the blende would have been attacked and the process would have been reversed, the blende forming the mat while galena and quartz were left behind. If dilute sulphuric acid, one part of acid to four of water, had been used, then both the blende and galena would have been attacked and if the ore had been finely pulverized, no ‘preferential’ separation would have resulted, both galena and blende finding their way into the float concentrate. However, with coarse material, the blende is much more highly charged than the galena and if the watch-glass be tapped and the contents given a vanning motion, the blende will gather most of the H2S bubbles and finally float, leaving galena and quartz behind. This shows that the electrification of minerals varies with different acids and also with different strengths of the same. This action of one mineral, ‘robbing’ the others of their bubbles, has not been utilized in practice, as yet, but there is no reason why ‘preferential’ separations could not be made on a large scale, utilizing this principle. Less air or gas would be necessary than in the present type of frothing-cells and a clean concentrate would be produced at once. A separation of blende, galena, pyrite, and quartz may be made as follows:

Add dilute sulphuric acid and pan off the blende; then add dilute nitric acid and pan off the galena; then add concentrated sulphuric or nitric acid, which attacks the pyrite so that it may be panned off.

Or the separation may be made with nitric acid alone, varying the strengths; with sulphuric acid, by use of the ‘robbing’ action described above or by use of hydrochloric and other reagents that attack one or another of the minerals more strongly than the others. If galena or blende and magnetite be treated with dilute sulphuric acid, the magnetite will not be acted upon by the acid, but some of the H2S bubbles generated by the sulphide will attach themselves to the magnetite, provided the bubble is formed near the magnetite. This illustrates the fact that electrical conductors in a conducting liquid attract electrified bubbles. A slight jar, however, will displace these bubbles; or a piece of sulphide in close proximity will rob the magnetite of the bubble, magnetite being a poor conductor.

Referring to the article on page 668 of the Mining and Scientific Press of October 30, 1915, describing a patent for preferential flotation of blende, galena, and pyrite, the second paragraph reads: “The new process consists of treating ores in a medium (i. e. sulphuric acid and sodium sulphite) that wets the zinc sulphide and which does not wet the lead sulphide or pyrite.” This phenomenon brings out nicely the part played in flotation by the “dielectric film.” When thio-sulphates, sulphites, or bi-sulphites are acted upon by sulphuric acid, there is more to the phenomenon than formation of S02 gas.

The following reactions take place when blende and galena are treated with sulphuric acid and sodium sulphide:

ZnS + PbS + 2H2S04 = ZnSO4 + PbSO4 + 2H2S
Na2SO3+ H2SO4= Na2SO4+ H2O + SO2
2H2S + SO2 = 2H20 + 3S

This sulphur thus formed is in a very fine state and acts as a dielectric film about the galena, for which it has a great attraction. Therefore, no frothing agent is needed in this case, as the dielectric film about the bubbles is formed by the sulphur similarly to the films of oil formed in the ordinary flotation processes. In the last paragraph of the above-mentioned article on ‘Preferential Flotation,’ the statement is made that “the procuring of the effect aimed at, is dependent upon the presence of a frothing agent, only when a reducing gas is introduced into the medium. It is not dependent on the presence of a frothing agent in the flotation medium, when a reducing gas is generated in the flotation medium by a reaction of a substance introduced into it.” In other words, if sulphur or any other ‘dielectric’ is liberated in a very fine state, by a “reaction of a substance introduced,” no frothing agent need be used.

This action may be nicely illustrated by taking 20 to 30-mesh galena and blende and treating them with dilute nitric acid on a watch-glass and observing the result under a microscope. The galena will gather all the H2S bubbles, when vanned. Now if the sulphuric acid is added and the watch-glass be tapped and the particles moved over one another, the H2S bubbles on the galena will be robbed by the blende. Sulphur may be seen surrounding the bubbles, the reaction being as follows:

ZnS + PbS + 2H2SO4 = ZnSO4 + PbSO4 + 2H2S
(1) H2S + H2SO4 = SO2 + 2H2O + S
(2) 2H2S + SO2 = 2H2O + 3S

In (1), the sulphur is formed from the decomposition of H2S; and would be charged oppositely to the sulphur formed by the decomposition of SO2 gas. In (2), we have both negatively and positively charged sulphur particles.

The larger bubbles, with sulphur particles adhering to them, may burst on reaching the surface and their film of sulphur will spread over the water. If particles of mineral had been attracted to the bubble, then these particles would have remained attached to the sulphur film, even after disruption of the bubble, showing the electrification of the dielectric film. The same phenomenon occurs when the film is a liquid dielectric, like oil.

The laboratory tests with chemical reagents generating H2S gas give the opposite results from the regular air-bubble flotation. The ordinary flotation method in practice is to add a little acid and frothing agent to the pulp. The sulphides are positively charged by friction, while the frothing agent and air are charged negatively. The oil surrounds the sulphides, but the film is so thin that the negatively-charged bubble is attracted by the positively-charged sulphide. If the film of oil is too thick, the attraction between the sulphide and bubble is too feeble, and flotation fails.

In the laboratory, H2S is charged positively, but the sulphides are charged negatively by chemical action. So the bubble attaches itself to the electrified sulphide.

In preferential flotation of galena and blende, in practice, the ore is treated with H2SO4 and Na2SO3. Sulphur is liberated and the galena is coated with the particles of sulphur positively-charged, while the negatively-charged sulphur coats the gas bubbles. The negatively-charged air bubbles of the flotation machine attach themselves to the positively-charged sulphur coating the galena and repel the negatively-charged blende.

In the laboratory experiments, the H2S bubbles are charged positively and are attracted to the negatively-charged blende and repelled by the positively-charged sulphur on the galena. In this case, the blende is floated. When H2S is blown into the pulp in practice, no sulphur is formed—the H2SO4 solution being too weak for this reaction. Therefore, to make a persistent froth, a frothing agent must be added to the pulp.

In conclusion, it may be well to call attention to the fact that for laboratory experiments in preferential flotation, any one of the sulphides may be separated from the other sulphides, (a) by the use of some reagent that attacks this particular sulphide and not the others, (b) by the use of a reagent that attacks one sulphide more vigorously than the others; in this case, the vanning motion allows the sulphide more highly charged to gather up the bubbles from the sulphides less highly charged, and if sufficient bubbles are collected, the mass of bubbles and sulphide will float. If not sufficiently buoyed, the mass remains submerged, but it is lighter than the other sulphides or gangue minerals and can be panned off or separated by hydraulic classification.

The second point of interest is the formation of a frothing agent, within the pulp, when reactions take place that liberate dielectric substances in a very fine state, electrically charged.

The third point is that laboratory experiments may not work out in practice, due to failure to understand the nature of the electrical charges of the bubbles, dielectrics, and particles of ore. A little stronger reagent or a different way of frictionally electrifying the bubbles and pulp, or too thick a film of dielectric or frothing agent causes the attraction to cease or change. It is 110 wonder that great difficulty has been experienced in the practical application of flotation to ores, when such delicate electric forces have to be considered.