Table of Contents
The present investigation set out to study the individual flotation behavior of leached quartz, unleached hematite, and unleached calcite with anionic and cationic collectors. The flotation response of mixtures of the minerals taken two at a time was then determined, and finally the experiments were repeated for mixtures of all three. Obviously there exists in nature no other quartz-calcite-hematite mixture like it so that, even if the authors could explain the behavior of their artificial system, it would be of little avail. The investigation did, however, produce some interesting and controversial results.
Materials and Experimental Procedure
Quartz, from the Hardin Mine, Dixon, New Mexico, was leached in concentrated HCl for 48 hours and then washed thoroughly with distilled water. Oolitic hematite, from Clinton, New York, containing Al, Mn, Mg, and Ca as its chief contaminants, and calcite, from Magdalena, New Mexico, containing Al, Fe, Mn, and Mg as prime impurities, were used without leaching.
All minerals were wet ground in a pebble mill and screened to give a minus 65, plus 150 mesh fraction for flotation tests and a minus 150 mesh fraction for electrokinetic studies. After drying at room temperature, magnetic material was removed and samples were stored dry in pyrex bottles. The minus 150-mesh material was hand ground in a Damonite mortar until the sample was reduced to minus five microns. The average diameter of these particles, measured in a Sub-Sieve Sizer, was found to be about 3.5 microns. Armac-TD, a high purity primary amine, and sodium oleate, freshly prepared every day, were used as collectors. All other chemicals were reagent grade.
Flotation tests were made in a miniature flotation cell using a one gram sample in distilled water. Solubility determinations of collector salts were carried out by means of a Nephelometer and a model 110 Turner Fluorometer. A Perkin-Elmer 421 Double-Beam Spectrophotometer was used for infrared studies, and electrokinetic measurements were made in an electrophoresis cell developed at the New Mexico Bureau of Mines.
Experimental Results and Discussion
Leached quartz gave no flotation at any pH or collector concentration.
The flotation behavior of calcite is shown in Figure 1. As the collector concentration increased from 2.8 x 10 -6 M to 2.8 x 10 -5 M, the recovery curves showed two peaks developing at pH 6 and 9, with a very pronounced trough in between. This trough coincided with the localized peak in the calcite recovery curve with amine collector (Figure 8). Further increase in collector concentration tended to even out the trough until at 10 -4 M, recovery was 100% from pH 6 to 12. Flotation at the pH 6.0 peak was always better and faster than at the pH 9 peak up to a concentration of 10 -4 M of collector.
The unusual shape of the calcite recovery curve suggested that two very different mechanisms might be responsible for flotation and a variety of tests were run to try to resolve the issue.
After mildly leaching calcite in a column with a pH 4.9 HCl solution, the recovery curve retained its shape and shifted slightly to lower pH values. Three different calcite samples, including a pure Iceland Spar variety, all gave similar recovery curves with a clearly defined trough. From these experiments it was apparent that surface contaminants were not responsible for the peaks or trough in the recovery curve.
Infrared studies seemed to offer a method of establishing the mechanism and tests were run on a Perkin-Elmer 421 Double-Beam Spectrophotometer. Calcite absorbs strongly for wavelengths between 6 and 8 microns, the region in which the carboxyl antisymmetrical group is located. Special techniques are necessary to avoid these peaks being masked in the extinction region, and to magnify them for identification. These techniques include reducing the calcite particles to sizes below that normally used, and increasing the concentration of sodium oleate. To obtain the spectrum of calcite at pH 6 it was necessary to acidify the pulp to about pH 3.5 and under these conditions calcite was extremely soluble, particularly at sizes of 1 or 2 microns. Calcium oleate immediately precipitated, and, apart from the difficulty of separating out the remaining calcite, it was impossible to distinguish between the precipitate and any compound formed at the calcite surface.
Electrophoretic measurements on calcite showed that the mineral was positively charged for all pH values in the range pH 5 – 11, in agreement with the results of Borisov. There are at least five ions resulting from the dissolution of calcite and since these ions and their hydrolysis products, together with H+ and OH-, will be potential determining, Parks is of the opinion that charge reversal brought about by a pH change is impossible. There seemed to be little merit, therefore, in trying to attribute one peak to physical adsorption and the other to chemisorption.
In the presence of collector, the equilibrium pH of calcite in distilled water was 8.7. Hence, by starting with a pH 3.0 solution and prolonging the conditioning time, flotation could be carried out at any intermediate pH. At a collector concentration of 2.8 x 10 -5 and no conditioning time recovery was 96% after a three minute float, and during this flotation interval the pH increased to approximately pH 6. The same test repeated with a one minute conditioning time gave 100% recovery in a one minute flotation time. All subsequent tests with prolonged conditioning gave 100% recovery with progressively shorter flotation times. These tests gave no real clue to the mechanism of flotation but showed that, after being contacted with a pH 3.0 solution, the calcite surface was very receptive to collector in free acid or ionic form. The reaction between HCl and calcite at pH 3.0 produces CO2, which, although not visible as gas bubbles, must be present on the surface of calcite. It seems that these gas bubbles, coupled with a surface cleaned by leaching, have a pronounced beneficial effect on flotation.
Several other tests, including flotation at elevated temperatures, flotation after washing the conditioned mineral addition of sulfate (slight depression) and fluoride (slight activation), failed to throw light on the adsorption mechanism and hence the following explanation can at best be speculative.
The solubility product of calcium oleate at pH 8 was found to be 1.5 x 10 -13. From thermodynamic data it is possible to determine the equilibrium concentration of the various ionic species present in a calcite-water system open to the atmosphere. Since sodium oleate is almost completely dissociated at this pH the concentration of oleate ions is approximately the same as the collector concentration. Calculations show that, for a collector concentration of 2.8 x 10 -5 M, calcium oleate should precipitate below pH 8.5 and as du Rietz suggests, the point of precipitation of collector salts maybe the critical condition for incipient flotation. He does not suggest by what mechanism flotation occurs under these conditions. Fuerstenau suggests that CaOH+ complexing with oleate ions to form Ca(OH (RCOO), either as an aqueous phase or as a precipitate, may be responsible for quartz activation at high pH. At pH 8, CaOH+ concentration is vanishingly small and the formation of such a complex in the case of calcite flotation does not seem probable.
As the pH is lowered the concentration of calcium ions increases and the concentration of oleate ions decreases, but at a slower rate than the calcium ions increase, so that, with the same 2.8 x 10 -5 M collector concentration, conditions are favorable for calcium oleate precipitation. However, free oleic acid is beginning to form by hydrolysis, and it is probable that physical adsorption of acid on the mineral surface is inhibiting flotation. At no stage in flotation was a precipitate visible.
The flotation peak at pH 9 may be attributed to the following reaction:
CaCO3 (surface) + 2RCOO- ↔ Ca(RCOO)2 (surface) + CO3=
As the pH is increased the concentration of CO3= increases and drives this reaction from right to left resulting in decreased flotation.
Unleached hematite responded readily to sodium oleate and gave good flotation in alkaline pulps, (Figure 2), when the mineral surface was negatively charged. This would seem to rule out an electrostatic model, which is the mechanism normally believed to be operative in the case of leached hematite. A recent investigation by Peck et al., on two unleached hematite specimens has shown that chemisorption of oleate reaches a maximum value at the isoelectric points (7.7 and 8.6) and flotation is also best at these pH values. They found oleic acid beginning to adsorb physically below pH 7 and they attributed the rapid fall in flotation recovery to this mechanism. Oko and Salmon attribute hematite flotation at pH 10.0 to an exchange mechanism between surface hydroxyl ions and oleate ions.
It should be noted that with a collector concentration of 7.4 x 10 -6 m and an Fe³+ concentration that is pH dependent, the solubility product of 4 x 10 -30 for ferric oleate was never exceeded. Hence precipitation of ferric oleate did not give rise to flotation. Reaction of oleate ions with ferric ions at the mineral surface was considered to be the mechanism which gave a hydrophobic surface and good flotation over the pH range 7-10.
As can be seen from Figure 3 the calcite behaved in much the same way when mixed with quartz as it did when floated alone. However, the trough was not so deep and the recovery peak at pH 9 was increased. If it is assumed that no collector is adsorbed on quartz, i.e., there is no quartz activation, than a collector concentration of 2.8 x 10 -5 M should give a better percentage calcite recovery for one half a gram than for one gram. The results showed this to be true in the range pH 6 – 10, but the peak at pH 6.0 was not affected.
It was expected that quartz would be activated by calcium ions at high pH through the formation of a calcium-hydroxy collector complex. Figure 3 shows that no such activation took place, presumably because at pH 10, the CaOH+ equilibrium concentration is only about 10 -10 M, too small a concentration to be effective. However, the activation of quartz, even though it does not lead to flotation, should result in some collector adsorption, and this, no doubt, limits the range of calcite flotation at higher pH values.
A mixture of calcite and hematite responded to flotation by sodium oleate as shown in Figure 4. Individually, calcite gave 100% recovery with a collector concentration of 10 -4 M at pH 8, and hematite gave approximately 20% recovery. Together they are mutually depressed. Mikhailova found that optimum flotation of hematite with sodium oleate occurred in the pH range 6 – 8, and calcium and magnesium ions depressed hematite. Although Gaudin, et al., found no metallic salt which improved the flotability of calcite with oleic acid, it is apparent from Figure 4 that calcite flotation was improved above pH 10 when mixed with hematite. For tests below pH 8 a brownish precipitate was again formed which obviously-had no inhibiting effect on calcite flotation at low pH values.
The flotation of hematite from a mixture of quartz and hematite (Figure 5) was almost unchanged from its flotation behavior alone. There was a smaller quantity of hematite present than when floated alone and, if quartz adsorbed no collector, the widening of the pH flotation range on the alkaline side might be expected. Manganese, magnesium and calcium ions present as impurities in the hematite could have activated quartz in alkaline solutions had their concentrations in solution been sufficiently high. Obviously they did not, and this suggests that activation of calcite in a calcite-hematite mixture did not stem from magnesium impurities in hematite.
When a mixture of all three of the minerals was floated with different concentrations of collector the results were as shown in Figure 6. The persistence of the calcite trough between pH 6 and 7 was not unexpected, but its narrowing as collector concentration increased was somewhat surprising. The complexity of such a system permits only a qualitative appraisal, but it is of interest to note that both quartz and hematite were activated slightly at higher pH values.
Quartz was readily floatable with amine at a low collector concentration over a wide pH range (Figure 7). Leached quartz was found to have an isoelectric point at pH 2.7 and was negatively charged above this value. Positively charged aminium ions are attracted and ionically bonded to the quartz surface to give good flotation. At low pH values flotation decreased due to the change in electrokinetic potential and competition from hydrogen ions. At high pH values recovery decreased as the free amine began to form from hydrolysis.
At a similar collector concentration of 2 x 10 -6 M, calcite showed slight flotation response at pH 6.5 (Figure 8) and had a very localized recovery peak at this same pH when the collector concentration was doubled. Flotation in this region was considered to be the result of an exchange between aminium ions and calcium ions whereby an amine salt was formed which was an effective calcite collector. As the pH was increased more carbonate ions were leached from the calcite and these combined with aminium ions to form alkyl carbamate, which was precipitated by cations in the pulp.
A collector concentration of 10 -4 M was necessary to induce flotation (Figure 9) and the recovery curve was markedly different from other published results for this system. At pH values above the isoelectric point (pH 6.8) the mineral was negatively charged and, as might be expected, positively charged aminium ions were attracted to the surface and gave a very localized recovery peak at pH 9.5. An ion exchange, accompanied possibly by the adsorption of free amine, is thought to account for the second localized peak at pH 6.5.
From the results of individual tests it might have been predicted that quartz could be separated from a mixture of quartz and calcite at either pH 5 or pH 8. Figure 10 shows that this was not the case. At pH 5 the concentration of calcium ions in solution is orders of magnitude greater than collector ions and these calcium ions must be competing with aminium ions for sites at the quartz surface. It was somewhat surprising that, even though calcium ions outnumbered collector ions a million to one at pH 6, quartz gave 100% recovery.
It is well known that cations in general are activators in conjunction with anionic collectors and depressants with cationic collectors. From Figure 11 it is apparent that quartz, which gave 100% flotation at pH 7 when floated with 2 x 10 -6 M collector gave only 80% recovery, with a tenfold increase in collector, when mixed with hematite. This could be due to Fe³+ occupying active sights on the quartz surface thereby preventing adsorption of collector cations. However, the concentration of Fe³+ is so small at pH 7 that this mechanism is not considered to be operative. More likely the hematite was tying up a significant amount of collector, leaving insufficient for complete flotation of quartz.
From Figure 12 it can be seen that calcite floated preferentially from hematite over the pH range 6-8. When mixed with hematite, calcite required a higher concentration of collector than it did when floated alone to give the same percentage recovery. Again hematite adsorbed some of the collector at the expense of calcite.
The results of floating a mixture composed of one third of a gram of each of the three minerals are shown in Figure 13. There was no flotation of any mineral at a collector concentration of 4 x 10 -6 M, but at 6 x 10 -6 M, calcite and quartz responded. Recovery of both minerals increased with collector concentration and at 1.25 x 10 -5 M, amine, both quartz and calcite gave 100% flotation. Although the peaks in calcite flotation always occurred at pH 6.5, the pH for maximum quartz flotation shifted to alkaline pulps as collector concentration increased.
The flotation behavior of leached quartz, unleached calcite, and unleached hematite with anionic and cationic collectors was determined with all possible combinations of the minerals.
Quartz did not respond to the anionic collector but gave good flotation with amine over the pH range for which quartz was negatively charged.
Calcite gave a very localized recovery peak with amine at pH 6.5 due to an exchange mechanism between calcium ions and/or hydrogen ions on the calcite surface, and aminium ions. With sodium oleate, calcite gave two distinct peaks on either side of pH 6.5 which were sensitive to collector concentration. Since the mineral had a positive zeta potential over the pH range studied, one peak could not be attributed to physical adsorption and the other to chemisorption. The first peak at pH 6 occurred in the pH region where calcium oleate was precipitated and flotation resulted, in some unexplained way, from this precipitation.
Flotation at the pH 9 peak was attributed to a chemical action between calcium ions at the surface and oleate to form calcium oleate.
Hematite floated well with oleate above pH 6.8, i.e., when its surface was negatively charged. The mechanism could be either ionic exchange or chemisorption. With amine, hematite gave two distinct recovery peaks. One at pH 9.5 was attributed to the attraction between positively charged aminium ions and a negatively charged surface. The lower peak at pH 6.5 was considered due to an exchange mechanism between protons at the mineral surface and aminium ions.
Mixtures of the minerals showed that, in the case of cationic collector, quartz and calcite floated in preference to hematite, and quartz floated in preference to calcite. However, the shapes of the recovery curves were not appreciably changed from those obtained for individual minerals.
With anionic collector, quartz did not float except when activated at high pH values, presumably by calcite. Calcite floated preferentially from hematite in all mixtures, but they were mutually depressed at certain pH values. The characteristic trough in the calcite recovery curve persisted when calcite was mixed with other minerals.