Why do Minerals Float

Why do Minerals Float

I was very much interested in reading an article by Charles T. Durell, appearing in the Mining and Scientific Press of September 18, under the caption ‘”Why do Minerals Float?’ However, I find myself unable to agree with Mr. Durell’s line of argument, and for the following reasons:

In the first place I believe that Mr. Durell has used loosely some rather obscure scientific terms which may cause unnecessary confusion to anyone not thoroughly familiar with the physical chemistry involved. The term ‘nascent gas’ is especially open to criticism. Doubtless Mr. Durell means the dissolved gas that can be liberated from the water, but that is hardly the ordinary sense of the term, and many modern physical chemists will object to the use of the word ‘nascent’ under any conditions whatever, or even refuse to recognize it, in spite of the old ideas that grew up around it. However, it may be that ‘nascent’ is a good term to use here in a figurative sense. Another term used by Mr. Durell and to which an objection might be raised is the word ‘occlusion’ as applied to the gases held by minerals. Mineralogists have long used this term rather loosely, but Mr. Durell does not seem to have taken it up in the same sense. As I understand it, there are three ways other than in visible openings by which gases can be held in solids; these are :

  1. In solid solution, in the same way that gases can be held in liquid solutions.
  2. In surface adsorption, where the layer of gas immediately in contact with the solid is found to be more or less tightly held by some force of attraction, unnamed, and is hence considerably compressed.
  3. In occluded form. This is a term the meaning of which has been much disputed; it is often used in the sense of one or the other of the terms above mentioned. Of late there has been a tendency to call occluded gas any gas that seemingly is held in some manner different from that designated by either of the other two terms, and in a manner not exactly understood. An example is found in certain substances that are known to be finely porous and that seem to hold gases in neither solid solution nor surface adsorption. Charcoal is such a substance. Possibly these have pores of such small diameter that they are of the same order of magnitude as the thickness of the adsorbed layer of gas, so that most of the gas held in them is present in a highly compressed condition. Charcoal absorbs so much CO2, for instance, that at ordinary temperatures the volume of CO2 held is enough to fill the known pore-space at a pressure sufficient to liquify it.

There doubtless are very fine openings in our crystalline sulphides, and they admittedly do hold some occluded gas, but good cases of occlusion have been found thus far only in amorphous substances like charcoal, hair, wool, glue, meerschaum, starch, etc. This fact tends to cast a doubt upon Mr. Durell’s hypothesis that the occluded gas in the flotative minerals plays an active part in the attachment of gas bubbles to the surface of the mineral.

Another physical fact that casts further doubt on this hypothesis is that occluded gases (and dissolved gases as well) are liberated from the substances occluding them only with difficulty; the last traces of them are still held even when the substance is placed in a high vacuum and heated to a considerable degree. It is as though the molecules of the gas were diffusing out through very small clogged pores. How this tightly-held gas could be liberated fast enough to compare with the exceedingly short time which it takes to accomplish flotation of a sulphide particle, is difficult to explain physically. Mr. Durell has further supposed that there is an osmotic travel of ions from the water solution, in which the ore pulp is made up, directly into the fine openings of the mineral particles. The surface of the particle acts as a septum and at the same time the molecules of the gas diffuse out and join the air from the solution, forming a bubble that floats the mineral. At the present time I cannot see my way to accept Mr. Durell’s supposition. Briefly stated, then, Mr. Durell’s hypothesis is that “nascent” gas from the water and “dissolved” gas from the solid meet and collect at the surface of the solid until a bubble large enough to lift the particle is formed, while the purpose of an oil is to form a persistent froth so that the particle will not be dropped back into the pulp. I am of the opinion that all the phenomena cited by Mr. Durell constitute no proof of any part of his hypothesis, but only furnish the basis for an inference that such could be the case. Whether it is possible to get flotation from water containing no dissolved gases and with minerals that have been treated for the removal of their occluded gases, no one knows. Possibly Mr. Durell is right when he says that “this phenomenon is worthy of investigation,” but, on the other hand, I believe that flotation can be explained more satisfactorily than by the conjecture that well crystallized minerals, such as the metallic sulphides, contain important amounts of either occluded or dissolved gases, and that they contain these gases in any greater amount than do the equally well crystallized gangue-minerals, such as quartz and feldspar. That surface films of adsorbed gas or air exist and may be of great importance, I firmly believe, but there is little evidence of any great difference in the amount of this gas on gangue and on flotative minerals.

As I understand Mr. Durell, the sole use of the oil in froth-flotation is the formation of a tougher liquid film around the bubbles of air, so that the resulting froth is more persistent. Mr. Rickard has pointed out that various writers have continually made the statement that the surface tension of water is increased by the addition of the oil, while as a matter of fact it is decreased. This fact Mr. Durell acknowledges, and yet he states that because the surface tension of the water is decreased, the tendency to float is likewise decreased. He implies that the reason is because the bubbles burst more easily. This seems strange, but in the absence of further data we can let it pass. My main comment is that while Mr. Durell believes that the only function of oil is the toughening of the surface film, we are not sure that such is the case. His hypothesis about “nascent” gas and “occluded” gas does not require the presence of oil; hence he has had to explain the use of oil or abandon the hypothesis.

Although I consider that Mr. Durell should be applauded for his courage in putting forward a hypothesis concerning a subject that has been of so empirical a nature up to date, nevertheless I have felt the necessity for challenging Mr. Durell’s hypothesis and of taking the liberty of advancing what seems to me to be a better one.

There are two such hypotheses which seem to be equally possible and it is not certain but that the two simply cover parts of a greater fact. One hypothesis can be stated in terms of the inter-facial tensions involved, and the other in terms of the electric charges residing on suspended particles.

The first hypothesis is based on some academic work done by Reinders. He based his work on some equations that Clerk Maxwell had derived from fundamental thermo-dynamic laws, leading up to a certain set of inequalities between inter-facial tensions of the phases involved. By ‘inter-facial tension’ I mean the state of strain existing at the zone of meeting of any two dissimilar phases. The surface of water in contact with air is under a strain which we call ‘surface tension,’ so that this surface acts like a tightly-stretched rubber membrane. Likewise the inter-face between a solid and water and between a solid and air or between oil and water is under a similar strain. Allusion has been made already to the surface adsorption of air on solids. The inter-facial tensions of many pairs of liquids are known, as well as the surface tensions of all manner of liquids in contact with all manner of gases. However, the inter-facial tensions of solids in contact with liquids have never been studied thoroughly on account of the difficulty of getting measurements that mean anything. Reinders deduced the following inequalities as applying to the case where a powder, or the particles of a colloid, is suspended in a liquid to which is added a second liquid that is immiscible with the first. Let us assume that the first liquid is water and the second oil, then expressing the inter-facial tension between the solid and water as T(S,W), the tension between water and oil as T(W,O), and the tension between the solid and oil as T(S,O), Reinders stated that:

If T(S,O) >T(W,O) + T(S,W) the solid powder will remain suspended in the water.
If T(W,S) > T(W,O) + T(O,S) the solid will leave the water and go into the layer of oil.
If T(W,O) > T(S,W) + T(S,O), or if none of the three tensions is greated than the sum of the other two, the solid particles will collect at the boundary between the oil and water.

It hardly needs to be said that here we find something very close to the conditions obtaining in the flotation process. In fact, the old Elmore bulk-oil flotation method fulfills exactly the conditions that Reinders had in mind. Below are given some tables of results obtained by both Reinders and Hoffmann in an experimental way, by suspending a definite powdered solid in water, adding a second immiscible liquid, and shaking. The letter w means that the powder remained in the water; the letter o means that the powder went into the oil layer; the letter s means that the powder went to the surface separating the oil from the water, and symbols like s(w) or s(o) mean that there was a good deal more of the powder in the inter-face than in the bracketed phase. Similar results were obtained with colloidal solutions.


Hoffmann worked a great deal with chemical precipitates and other artificially prepared products. However, the laboratory method involved ought to be good in the study of flotation processes for a possible method of floating oxidized minerals. Then it might be possible to convert a successful bulk-oil process into a frothing process in the same way as the old Elmore bulk-oil method went through the stage of granulation and classification to a final frothing process such as we


see today. I have done some work along these lines and have planned considerable research work in the same direction. Meantime these notes might as well be available to others in suggesting lines of useful research. The comments to be made on the above tables are without end, but the tables are given here mainly to suggest the possibilities of further work.

The question arises as to whether it might not be possible to apply a set of inequalities such as those of Reinders, or even to apply Reinders’ inequalities direct, in the prediction of results for froth flotation. In froth flotation we have at least four phases—solid, water, oil, and gas—unless the oil happens to be soluble in water, in which case we are reduced to a solid, a solution, and a gas. We have interfacial tensions between each two of the phases, making six tensions altogether, and mathematical expressions covering such a case must necessarily be much more complex and exhibit a greater number of possibilities. The problem is more difficult, but it should be capable of solution. Fig. 35 shows a fanciful magnification of one possible arrangement of the particles of solid, droplets of oil. and spherules of air, in the liquid of the ore-pulp, being subjected to flotation, at the moment when a bubble begins to raise a particle of mineral to the surface.Air

Fig. 36 shows a possible way of applying Reinders’ inequalities direct without any modification. It is assumed that the oil forms an envelope on the inner surface of the air bubble so that the air is nowhere in contact with water. Mr. Rickard has called our attention to the work of Devaux, published in the annual report of the Smithsonian Institute, in which it was found that a droplet of an oil when placed upon a plane surface of water will spread, of its own weight, until it forms a film only one or two molecules thick. This fact allows an explanation of how the small amount of oil used in the froth- flotation processes could be so efficient. If it so happens that the oil could coat the inner surface of an air bubble, the powdered mineral would be able to take up a position on the inter-face between the water and the oil without any reference to the air in the bubble.Bubble

By reason of the known low adhesiveness of oil and water it is doubtful if the air bubbles could be completely mantled by oil, as the oil would be too liable of its own weight to slide down to the bottom of the bubble to the position indicated in Fig. 37. Even here, the oil could carry mineral on its water inter-face (in case the oil and water do actually get into contact) and Reinders’ criteria would still apply. In a Callow flotation machine having glass sides it is sometimes possible to see particles in just such a position. However, this case does not prove that the top side of the bubble is coated with oil or mineral, while the bubbles of a mineral froth on top of the pulp are seen to be covered completely with particles of mineral.

If the mineral tends to enter the oil phase completely and leave the water, the mineral grains present only an oil surface and in case oil droplets tend to collect at the inter-face between water and air (by Reinders’ criteria) we could have the case illustrated in Fig. 38.


This case, as well as the one illustrated in Fig. 36, would allow of the air bubbles becoming completely covered with oiled mineral.

Other phases are possible, but these are given to show how very feasible it is to get an explanation of flotation in terms of inter-facial tensions.

In an investigation conducted by the Minerals Separation, the ‘contact angle’ of various minerals with water was examined to find at what angle the mineral had to come in contact with a water surface before it was wetted and could sink. A glance at Clerk Maxwell’s famous paper on ‘Capillarity,’ upon which Reinders’ work is based, will suggest immediately the explanation of a contact angle, and that it is the result of a certain equilibrium of inter-facial tensions of air, water, and solid. Valentiner has likewise investigated the contact angle and its hysteresis under certain conditions and has connected it very definitely with capillary phenomena. There can be no doubt that there is a close parallelism between the angle of hysteresis of the contact angle and the ability of a mineral to float. But if we go no further than to observe the parallelism we cannot designate the statement of the parallelism as a theory, although we might be able to predict by its means whether a mineral would float.

To go into this a little farther, and indeed along the line suggested by Mr. Durell, we ought to consider the properties of the surface layers of the substances involved. For example, the plane surface of water in contact with air is known to have considerably different properties from the inner bulk of the water. BulkIn Fig. 39 the film is shown magnified in thickness. It acts like a tightly stretched elastic skin, due to what we have long called a ‘surface tension’ of 81 dynes per centimetre, as is usually given in text-books. (This means that for a strip of the surface film one centimetre wide, a longitudinal tension of 81 dynes has been measured at ordinary temperatures, and there is a definite tension for each temperature.) This tension of the surface film is one of its most commonly known properties, but some other interesting points about it are given in the following:

Its thickness, varying with temperature and other conditions, has been estimated to be all the way from 4 X 105 to 108 cm. Its density averages 2.14 as compared with 1 for bulk water, although it is doubtless more dense at the immediate surface next to the air and gradually shades off into that of bulk water. This consideration probably explains the wide variation in the results of the measurement of the thickness, as one method might be less delicate than another and hence not take account of some of the layers of the film that are nearly bulk water in their properties. This average density, however, is illustrative of the magnitude of the force involved because water is a substance that resists compression, and it has been calculated from the known compressibility of water that the force necessary to compress it to twice its ordinary density is some thousands of atmospheres. Such a compression should liberate heat, and, in fact, the heat liberated when a definite area of new surface film is formed has been measured and found to be 0.00315 cal. per sq. cm. Being so highly compressed, its specific heat might be expected to be different from that of bulk water and has been measured as being nearly 0.45 instead of 1. This low specific heat approximates that of ice.

One important property of this film is that it will often take up dissolved substances in different proportion from the amounts in which they are taken up in the bulk solution, and there is always a definite equilibrium between the two. This is known as surface concentration or ‘surface adsorption,’ and has been dealt with mostly in colloid chemistry, where the large amount of surface of the finely divided solids is large in comparison with their weight. In case a greater proportion of the substance is concentrated into the film than there is into the bulk water we have positive adsorption; and in the reverse case, negative adsorption. The properties of these inter-facial films have been found to be greatly modified by small amounts of dissolved substances and the properties of colloids are hence likewise greatly changed The importance of the study of inter-facial films becomes obvious.

Finally, there is a most important fact about the film of water in contact with air. It has been found that there is a difference of potential of 0.055 volts between the two surfaces of the film. The density of the static electric charge at this potential is 4 X 105 coulombs per sq. cm. This electric charge is markedly influenced by electrolytes in solution and can be increased or decreased, even passing through zero and then increasing in the opposite sign. All interfacial films have likewise been found to be charged in one way or other. Industrial applications of this fact are legion. All the technical handling of clays is now conditioned by the use of electrolytes in this manner, and the question of emulsions of all kinds of oils in water is closely bound up with it. Cottrell’s precipitation process of suspended particles of solids or liquids in gases does not escape these considerations. Small particles of solids, liquids, or gases suspended in either liquid or gaseous media are found commonly to carry electric charges, due to various combinations of factors which affect the double electric layer of the inter-facial films.

Since the size of many of the particles of minerals treated by flotation is of the same magnitude as that of many colloids we cannot escape from calling ore-slime “coarse suspension colloids,” and must apply all the laws of colloid chemistry to our problem.

The electric charges on suspended particles allow another possible explanation of flotation phenomena. We find in some of the colloid chemical literature that quartz particles when suspended in water are negatively charged, pyrite particles positively charged, oil droplets are negatively charged, and air bubbles negatively charged.15 The charges are somewhat small compared with the weight of the particles, so that they are hardly strong enough to cause negatively-charged quartz to stick to positively-charged pyrite, as they can have only a few points of contact, and currents in the water could easily tear them apart. However, the negatively-charged droplet of oil, which is repelled from a negatively-charged particle of quartz, can wrap itself around the positively-charged pyrite particle so that they will stick together, and the same applies to air bubbles. The other sulphides known to be flotative have positive charges when suspended in water or can be made to assume positive charges by the use of the proper amount of the proper electrolyte. So it can be seen that the application of these principles gives no difficulty in explaining flotation from an entirely new standpoint. The large effect of a small amount of sulphuric acid on the conditions of flotation does not seem strange at all in this light, and we do not have to retreat to the purely imaginary supposition that osmotic pressure is acting through the surface of the mineral, as does Mr. Durell.

The inter-facial tension and the charge on the inter-facial film are two different physical properties of one and the same thing. I have shown that an appeal to either property is enough to build up a working picture of flotation phenomena that is simpler and more probable than that of Mr. Durell. I do not know how much there is in his contention that air bubbles will not attach themselves directly to the particles, or that only the dissolved air can thus attach itself; it may be that this is all correct, without interfering with the explanation that I have put forward. However, I hesitate to accept such a conception. The underlying cause of the tensions and of the electric charges is the same thing—some strange molecular, atomic, or other force manifested in ‘adhesion,’ ‘cohesion,’ or even ‘gravitation,’ if you please. No one can claim that electric charges carry the whole explanation of flotation, nor can it be stated that it is merely a question of a balancing of inter-facial tensions. Both will doubtless have to be considered.

Although much more could be said on the subject, I have only attempted to point out that there are certain scientific principles that can be applied to our problem, with great chances of success, in bringing us nearer to a definite understanding of flotation. Physical chemistry has been a recognized tool of metallurgists for some time, although little used by most of them, and now a particular branch of physical chemistry—colloid chemistry—is beckoning to us alluringly. All questions of the treatment of ore-slime should be studied in this new light. The results of an application of this idea in our own laboratory have been astonishing, and we hope that we may soon be able to publish them.