US2120485A - Ore flotation - Google Patents

Description

Patented June 14, 1938 NlTED SI ORE IFLGTATION Julius Bruce (Hemmer and Robert Gibson OMeara, Rolla, Mo., assignors of one-fourth to Semmes and Semmes, Washington, D. 0., a firm composed of Harry H. Semmes and S.

Warwick Keegin No Drawing. Application December 29, 1933, Serial No. 704,508

Claims.

This invention relates to the beneflciation of ores, and more particularly has reference to the flotation of tungsten. It will be seen, however,

that the invention embraces the flotation of minerals other than tungsten from a gangue composed principally of calcite or other carbonate minerals.

The principal tungsten ores scheelite (CaO.WO3), ferberite (FeO.WO3), Wolframite (Fe.MnO.WO3), and hubnerite (MnO.WOz)are usually complex and contain a variety of associated minerals. The most common gangues are quartz, feldspar, garnet, mica, sulphides,v arsenides, carbonates such as calcite, ankerite, etc., and phosphates such as apatite. Particularly to be noted in this connection is the fact that the contact-metamorphic scheelite ores, which are the most important, usually contain a large amount of calcite or other carbonate minerals.

'Heretofore, tungsten ores have been beneficiated by hand-picking and gravity concentration processes, such as jigging and tabling. In order to achieve any semblance of success by gravity concentration processes, however, there are two conditions necessary. In the first place, there must be a pronounced difference between the specific gravity of the mineral and the specific gravity of the gangue, and in the second place, the mineral and the gangue cannot be too finely interlocked, but must be susceptible of separation by relatively coarse crushing. In the absence of these two conditions, the concentrates are low grade and must be retreated by expensive chemical or pyrometallurgical processes in ever,- even when the two conditions are present,

the concentrates, while of marketable; grade, nevertheless account for a low recovery of the tungsten. Furthermore, no satisfactory method for handling the fines has heretofore been available.

At the present time, only the high grade and less complex ores are worked, and even in such case there are tailing dumps which contain large quantities of recoverable tungsten, provided a suitable method of beneficiation can be devised.

The market requirements for tungsten ore are relatively rigid. Concentrates are sold by the short ton unit of contained tungstic oxide (W03) and the base is usually taken at 60%. Higher grade concentrates bring a correspondingly high- .er price, and often a premium is given for exceptionally purematerial. Concentrates of lower grade than 60% are dificult to sell and usually a penalty of several dollars a unit is imposed. In addition to the required tungsten content, a concentrate for the ferrotungstic market must not contain more than 0.05% of copper, bismuth, antimony, tin, arsenic, sulphur, or phosphorus. Although some of the impure concentrates are used for chemicals, they have a relatively small market.

Thus it is apparent that a process such as flotation, which would successfully recover the fine mineral now wasted and enable high grade con-'- centrates to be produced from the more difllcult ores, would be of value in the concentration of tungsten ores. Flotation would serve as a useful adjunct -to gravity concentration processes by making the milling circuit more flexible and permitting a higher recovery of an improved grade concentrate. By providing a suitable flotation process, the impurities could be more thoroughly rejected and it would be feasible to mill ores which it is now considered impossible to treat.

While attempts have heretofore been made to float tungsten, no process has yet' 'been devised whereby a high grade and merchantable concentrate can be obtained The difllculties in the past have resided, not so much in the inability to float the tungsten, as in effecting a separation between the minerals and the various gangues with which they are found.

As stated above, our invention also comprises the flotation of various minerals (particularly fluorspar, barite and scheelite) from calcite and other calcareous minerals. The separation of these minerals by flotation methods has heretofore been considered practically impossible. If it be a fact that flotation by a fatty acid or derivative thereof depends upon the formation of a relatively insoluble soap at the mineral surface, then, on account of the common calcium ion, fluorspar, scheelite and calcite should act the same in flotation, and barite should be similar in its fiotative properties. This idea is supported by the results obtained by many investigators working on a variety of ores. Separation by flotation has been attempted by controlling the reactivity of the collector so that reaction pro-' ceeds with only one mineral; these attempts have been made by controlling the pH and also by using various fatty acids of relatively short hydrocarbon chains. However, all previous attempts to separate these minerals from calcite or other carbonate mineral by flotation, have failed.

To overcome the above mentioned difliculties is one of the objects of our invention.

Another object of our invention is to devise a process, or processes, for the flotation of tungsten.

Yet another object of our invention is to devise a flotation process for tungsten ores in which the concentrate is relatively free from impurities, and is of a high and merchantable grade.

Still another object of our invention is to devise a process for the flotation of certain minerals from a gangue in 'which calcite, or other car-' bonate mineral, is an important constituent.

- To accomplish the above, and other important objects, as will be hereinafter observed, our invention, in general, comprises the means and steps herein recited, it being distinctly understood, however, that various changes, and modifications, may be made therein without departing from the spirit of our invention or exceeding the scope of the appended claims.

The results herein set forth were obtained only after lengthy and numerous experiments, and the specific data submitted herewith is illustrative of the success we eventually obtained.

Scheelite ore The first tungsten flotation tests made by us were on a scheelite ore from Nevada in which the gangue was predominantly quartz with a small amount of apatite. The early tests indicated that the scheelite was more fioatable than the quartz, and we found that the use of a fatty acid or a fatty acid soap, alone or together with a frother, and an alkaline reagent,-'gave -a satisfactory separation of scheelite. Any of the well known fatty acids or fatty acid soaps may be used, and the frother may be of conventional character, such as pine oil, cresylic acid, the alcohols, or any of the moderately soluble heteropolar organic compounds which lower the surface tension of water. Caustic soda, soda ash or sodium silicate may be used as the alkaline reagent.

In one typical test, such an ore, assaying 2.38% tungstic oxide, gave a. concentrate of 62.72% W03, which constituted 89.7% of the total W03 content of the heads. The rougher concentrate was cleaned twice to give the final concentrate, and in commercial operation the middlings, which assayed 1.03% W03, would be returned to the flotation circuit, and the mineral would be recovered without a serious sacrifice in the grade of the final product.

, The ore was ground to -l00 mesh. As in the case of all the tests reported herein (with one exception, which will be referred to), the pulp contained 17% solids. Flotation was accomplished in a mechanically agitated machine of' standard design. The reagents used per ton of crude ore were as follows:

Rougher Cleaner Pine oil 0. 12 Oleic acid l6 Sodium oleate .40 Sodium carbonate Sodium silicate 40 rougher or cleaner. The acids that can be used.

to depress the apatite may be the inorganic acids, such as hydrochloric, nitric, sulphuric or sulphurous, and the short-chained organic acids such as formic or acetic. A small amount of acid sumces, merely enough to make the pulp neutral or slightly acid. In no case should the acid be added in an amount suflicient to dissolve the gangue.

. A fatty acid or derivative, or both, may be used 6 a as the collector, and, as in certain other instances herein, a frother is not necessary under some circumstances. A small amount of a soluble sili- 4 cate may be used to depress the quartz. The silicate maysometimes be omitted, however. The acid may be used either in the rougher or the cleaner, or both, but we have found that a slightly higher recovery of tungsten results when it is used in only the cleaner.

In a typical test on the Nevada scheelite ore having a gangue composed of quartz and apatite, and the heads assaying 3.16% WOa and 0.040% P, a high grade concentrate was produced assaying 77.06% W03, .036% P. and .03% S. This concentrate accounted for 87.6% of the total W0:

- and only 3.3% of the total P. A mechanically Cleaners Rougher Formic acid It will be observed that the concentrates produced by this process were of a high grade and that the phosphorus and sulphur contents are well within the limits prescribed by the industry. We found that the rougher concentratescleaned well.

In neither of the preceding experiments did the ore contain any sulphide minerals. In view of the fact that the flotation reagents employed for sulphides sometimes have a deleterious efiect on the subsequent flotation of oxides, we conducted a series of experiments on a sample-of Montana scheelite ore, in which the scheelite was associated with gold, arsenopyrite, mica and quartz. This ore also contained 0.636% of phosphorus as apatite. a

We first conditioned the ore and'then proceeded to float the gold and arsenopyrite. Flotation of the sulphides was followed byflotation of the scheelite in a rougher and three cleaning stages. High grade tungsten concentrates which were low in phosphorus and arsenic, were obtained, while the arsenopyrite concentrates contained the'gold and only a small amount of tungsten. The results of a typical test on this ore are as follows:

Sulphide concentrates... 5. scheelite concentrates 4 Middlings This test indicates that selective flotation canbe successfully applied to tungsten ores.

The ore was ground to mesh and a me- 7 chanically agitated flotation machine was used. The ore was first conditioned by the use'of sulphuric and lactic acids. Potassium amyl xanthate and hardwood creosote were employed for the flotation of the sulphides. under most conditions, other conventional sulphide flotation reagents might be substituted. An additional amount of lactic acid was added to the sulphide rougher to depress the mica and allow a clean sulphide concentrate to be produced.

After the flotation of the sulphides, the scheelite was floated with oleic acid and sodium oleate, and a small amount of silicate was added to retard the quartz. These rougher concentrates were cleaned with sodium silicate to depress the quartz and mica, and a final cleaner with formic acid depressed the remaining mica and apatite. It will of course be appreciated that suitable substitutions may be made for the above mentioned reagents, as occasion may warrant. The amounts of reagents were as follows:

' sulphide scheelite Cond. Rougher Rougher 01.1 01.2 01.3

Sulphuric acid Lactic acid K amyl xanthate. Hardwood creo- Sodium oleate. Sodium silicate. Formic acid .i

We have found that the arsenopyrite readily floats in the acid circuit, and that the lactic acid is especially effective in depressing mica both in the sulphide flotation and in the flotation of scheelite. It is also effective in depressing apatite but to a lesser degree than the inorganic or shortchained organic acids.

As will be seen from the above. the tungsten concentrates in such a selective flotation process are of a high and marketable grade, and of a low phosphorus and arsenic content.

The three foregoing tests were made on scheelite ores containing a siliceous gangue, and show that the scheelite can be concentrated by using a fatty acid or a derivative soap, or both. When there is no phosphorus present in the ore, an alkaline circuit may be employed, but where phosphorus is present, an acid or non-alkaline circuit is necessary in order to secure lowphosphorus concentrates. All of these ores were free of the calcareous minerals which occur abundantly in the contact-metamorphic ores.

Contact-metamorphorzc scheelite ores The presence of calcite or dolomite in scheelite ores has heretofore been considered as preventing the separation of scheelite by flotation. The

reagents that have previously been used have not been effective in retarding or depressing the alcite and inasmuch as calcite, scheelite and apatite have about the same floatability, a separation between them has been considered impossible.

We have now devised a process by which scheelite can be successfully concentrated from ores containing 'a calcareous gangue, and we have been able to secure high grade concentrates, low in impurities. We have conducted our experiments on a large number oi ores varyingconsid- We believe that i to be varied to meet the need of the ore, and that ordinarily the proper amounts can only be determined by experimentation. As a general rule, however, a greater amount of calcareous gangue requires a greater amount of the metal saltsilicate referred to hereafter.

Referring to one specific experiment, carried out on a Nevadacontact-metamorphic scheelite ore, the ore was ground to 65 mesh and was floated in a mechanically agitated flotation machine. Inasmuch as this ore contained pyrite and sphalerite, the sulphides were first separated by the use of potassium amyl xanthate and aerofloat, althdugh any of the other conventional sulphide flotation reagents could be used. Inasmuch as it is necessary to thoroughly remove the sulphides in order to produce a subsequent tungsten concentrate sufficiently low in sulphur, a relatively large amount of reagent is required. We have found that the sulphides can be' floated in an acid, neutral or alkaline pulp, without impairing the floatability of the tungsten.

After removal of the sulphides, the remaining ore is conditioned for a short time with a heavy metal salt, such as copper sulphate. We have found that the salts of other metals, such as silver, mercury, copper, lead, zinc, or manganese, which form relatively insoluble carbonates, can be employed. While in the specific test to which we refer, copper sulphate was used, other salts of copper such as copper nitrate, or copper ammonium hydroxide, are as effective. We have referred to introducing the metal salt in the conditioning stage, and prior to the flotation of the scheelite, but we have found that it may also be added to the sulphide conditioner orduring the sulphide flotation. if the particular salt has no retarding effect on the sulphides. Of course, if it has such a retarding effect on the sulphides, it must be added after the sulphides have been floated.

After proper conditioning, a soluble silicate,

such as sodium silicate, is added and thoroughly v mixed in the pulp. After the silicate has been sufiiciently mixed, a fatty acid or its derivative soap may be addedto float the scheelite. The

by using an acidic constituent such as acid dichromate, chromium nitrate, or any of the inorganic or short-chained organic acids. In

general, we have found that in tests in which the proper amount of metal salt and soluble silicate have been used in therougher, silicate alone will readily clean the rougher concentrate. If the proper amount of these reagents have not been used, however, then a final cleaning in a neutral or slightly acid pulp is necessary to depress the remaining calcareous material and apatite.

In the specific test referred toabove, the following reagents were employed in the amounts indicated:

good recovery of mineral in a concentrate of marketable grade.

.. In a typical test, on a ferberite-quartz ore containing an appreciable amount of ankerite, and the heads of which assayed 103% woo, a concentrate was obtained having a W03 content of 59.54%, which accounted for 79.1% of the total W03. Pounds perton oi ude ore This ore was ground to 200 mesh, and the Reagents flotation was carried out in a mechanically agi- Sul- Condi- Rough Cleaner 1 Clem tated flotation machine. The ore was passed PhidB through a rougher and three cleaners. The reagents used, and their amounts, were as follows: Kain lxanthate 0.40 ;';oi mi? "is 33:11:; Cleaners lllllS 08 8---- l Rougher i .96 odiu: leate 20 1 2 3 As indicated above, the concentrates were of a Crggylic acid 0.0a 0.08 0.08

high grade and were low in their phosphorus and $3 3 g g sulphur contents. The results of this specific Sodium n dl-oiidoI I50 I test were as ggigt'ts'gssig; .08 .16

A t Wt. my Ween Pemnm It will of course be appreciated that certain 20 pct. total W0; substitutions may be made in the reagents listed P S in this application without materially affecting the results obtained. sulphide concentrate &0 2,74 7,0 with respect to the sulphuric aci used in the 5 ggh Mamie flnal two cleaners, we found that the ankerite and lddlin .8 14.64 20.3 Taillngs .4 .60 4.5 siliceous gangue were effectively depressed. In $2332 3-3 337 100-0 other tests on ferberite ores, an acid was used in "'7 the rougher as well as the three cleaners, arid As will be seen sodium Silicate was the only we have found that acid dichromate (iis vlery alkaline reagent employed, and it was used in {active in depressing the gangue an so 6 30 the scheelite rougher. We have found as a result apatite' In one test g i' s; fi f of'other tests, however, that the use in the rougher. dichromate was use at t i or cleaner of caustic soda or soda ash inaddition of crude one in the muglher 2 a per i to the silicate gives -improved. results. The reeach of the three, clean n; 5 ages, concen ra e covery is increased without a serious sacrifice in of W03 was recovereg g g g 35 the grade of concentrate. In a recent test, 86% product? m mm treating a w of the mineral was recovered in a concentrate asproduct had a composition of 3 e saying 70% tungstic oxide concentrate accounted for 69.8% of the total W03,

We have also found that scheelite ores free of and had a Phosphorus nt of only 022 %d8.l1ld

40 calcite, but containing considerable apatite may 945% of sulphur 231" i? 5 40 also be effectively concentrated by using the metal h test were 0181 acldf s um 0 ea 8 an t u y salt-silicate reagent combination in conjunction 'alcohoL m 1 1 with the fatty acid. From such an ore, tungsten In certain ferbeflte ores 6 5 parconcentrates low in phosphorus can be produced I uqularly sensitive to depression by acid and in by cleaning with a soluble silicate r acidic such cases we have found that the introduction of 45 reagent. a small amount of a metal salt renders the min- Ferberite ores eral less sensitive but does not impair the eihcacy v of the gangue-depressing property of the acid. The flotation of ferberite has heretofore been Although not wishing t be bound by our exnlana- 60 considered impossible. Our investigations have mm of the theory we beneve t t th metal salt ihown h i was unwarranted although functions to form a comparatively insoluble metal ferberite is t as floatable s scheelite, we h oleate coating around the'particles, and thus renddevised a flotation process for the successful ree175 t particle 1 itive to depression by the covery of'ferberite. acid. Regardless of the theory, however, the net 56 Most ferberite ores have a siliceous gangue result higher ov r 'of mineral in a conwith only small amounts of sulphi s n p centrate of slightly better grade. This metal While the Ores which are liberated at m e salt-acid combination also serves to efiectively de-' may be readily concentrated by gravity processes, press the apatite. It should be stated, however,

' ferbeh iite s,d f t l fl 96 mm h ggiufi t that ba precise control of the acid will give comoo crus ngpro ucesrc nesw c are c o paraleresul, to rec by process other t flotation Ferberite seems to be best floated in an acid cirqlcft y. t n as 881881191 3:311:21 to cult, and, notwithstanding the fact that such ores 8m 11 Pro 8 n of e 1 Or are relatively free of carbonates, any apatite contain an appreciable amount of homstone comwhich may be present wouldbe depressed by the posed of minute grains of lerberite disseminated slight acidity of the circuit. 05

in a matrix of chalcedonic quartz. Inasmuch as During the course of the experimental work on such ores require line grinding for liberation, they the flotation of ferberite ore, some time was spent can be successfully treated only by a process in the "Boulder tungsten district of Colorado. adapted to handle material of this size, and we Flotation of ierberite under mill conditions was have found that flotation of these ores gives a attained. However, several difllculties had to be 70 overcome which are rarely encountered in laboratory work. The low temperature (34 F.) of the water available made flotation diflicult. The ground ore was so dispersed that although the coarlo mineral was recovered by flotation in the usual manner, the slime losses were too great. Further, the consumption of fatty acid or soap was greatly increased and flotation was slow. Coagulation of the pulp by customary settling reagents such as lime or sulphuric acid further increased the soap consumption and did not improve tl'ie flotation. It was found, however, that the use of an alkaline reagent, as soda ash, or a salt such as calcium sulphate, followed by a small amount of an acid gave selective coagulation of the mineral, whereas the flne gangue remained dispersed. The reverse order can also be used acid followed by a base. The coagulated mineral was then readily floatable with the customary reagents and enabled a good separation to be obtained with low tailing losses. Emulsifying the fatty acid in the soap prior to flotation materially reduced'the quantity of reagent necessary and gave normal flotation in spite of the cold temperature of the water.

Wolframite Cleaners Rongher Oleic acid .96 Sodium oleute Acid dichroinate Butyl alcohol .I

It will be observed from the above that the rougher concentrates were cleaned once with no reagents. We found that the wolframite floated readily and that the dichromate effectively depressed the gangue.

Hulmerit'e Due to the higher manganese content of hubnerite, this ore is even more floatable than wolframite, and the same reagents employed on ferberite or wolframite may be used for the flotation of hubnerite. A typical test on an Arizona hubnerite ore which assayed 4.20% W03 and 0.05% P resulted in the recovery of a concentrate assaying 71.56% W03, .012% P and 043% S, accounting for 90.2% of the total W03 content of the ore.

This ore contained pyrite and chalco-pyrite, which were first floated by the use of 0.25 lb. of potassium amyl xanthate per ton of crude ore and .14 lb. of butyl alcohol as a frother. Sodium silicate at the rate of .65 lb. was used to prevent the tungsten from floating with the sulphides. We found that the use of these reagents did not impair the subsequent floatability of the hubnerite. We also found that butyl alcohol is preferable as a frothing agent, inasmuch as it secured an equally good recovery of sulphides but at the same time did not float as much tungsten as certain other frothers, such as pine oil.

In the hubnerite rougher, oleic acid at the rate of 0.401b. per ton of crude ore and sodium Complex ores From the above, it will be noted that certain of the tungsten minerals are more floatable than others, and we have found that if they are arranged in their order of increasing floatability, such order would be: ferberite, wolframite, hubnerite and scheelite. We have found that scheelite can be successfully separated from Wolframite, and, although this would not ordinarily be done, it might be necessary to effect such a separation in order to produce marketable grade products from an ore that contains some other.

additional mineral, such as fluorspar. In such event, fluorspar and scheelite could be concentrated into one product and the wolframite into another separate product. The separation of the fluorspar could be effected in a retreatment of the fluorspar-scheelite product. In the absence of such a procedure, and with the present state of the art, separation would be almost impossible if all three minerals were floated together.

Our experiments in this connection were conducted on an ore from the Springvale mining dis-' trict of Washington which assayed 1.48% tungstic oxide and 0.009% phosphorus. Wolframite -and scheelite constituted the tungsten and were associated with sulphides and gangue composed mainly of quartz and mica, and there were also present fluorspar, calcite, ankerite, apatite and tourmaline. The complex nature of the ore and the low tungsten content made testing diiflcult, but the results of several tests showed that the ore could be satisfactorily concentrated by flotation.

In a typical test, the results were as follows:

W t Assay, per cent P f g ct. 0 Product Pct. total W0 W03 P S sulphide concentrate 1.9 2. 09 2. 7 Fluorspar concentrate"... 2. 4 1.53 2. 5 scheelite concentrate 5 59. 85 20. 2 Woliramite concentmt 9 08. 32 020 055 41.5 Composite milldlings 5. 6 8. 44 32.0 Tailings 88. 7 .02 l. 1 Composite 100.0 1. 48 mo. 0

The ore was ground to 150 mesh, and floated in a mechanically agitated flotation machine. The procedure was, first, to float the sulphides, and thento float the fluorspar, scheelite and calcareous materials in a second concentrate; and-finally, a concentrate of the wolframite is secured. The sulphides were floated with xanthate and wood creosote, although other conventional sulphide reagents might be employed, and a sufiicient quantity of a soluble silicate is included to temporarily depress the scheelite and wolframite.

The fluorspar, scheelite and calcareous materials were recovered in a concentrate that was relatively free of wolframite, and a small amount of olelc acid or oleate is all that is required for this step, if suflicient silicate is employed in the sulphide rougher. I If sufllcient silicate is not used in the sulphide rougher, it may be necessary to use it in the scheelite-fluorspar rougher. In any event, a silicate is used in the first cleaner in order to depress any wolframite, quartz or mica. The middling product was combined-with the wolframite rougher concentrate in order to recover the wolframite. The fluorspar-scheelite concentrate was then recleaned with formic and hydrochloric acid in order to'effectively depress the scheelite and yield a concentrate of fluorspar that is fairly free of scheelite. The middlings are enriched in scheelite and are of marketable grade with respect to the W: content.

The tailings from the scheelite-fluorsparrougher stage are subjected to a flotation treatment in which oleic acid and a small amount of oleate are employed to float the wolframite. The concentrate thus obtained is a medium grade and is contaminated with mica, quartz and some ankersents quantities of reagents in the above mentioned test:

Clean- Cleaners ers Scheellte Woliram rougher rougher Reagents xanthate..-

As above stated, the flnal concentrates assayed well over 65% W03, and were low in impurities. The results clearly show the feasibility of treating this type of ore by flotation.

The results of all the foregoing tests show conclusively that tungsten ores can be successfully treated by flotation methods to yield high grade concentrates that are low in impurities. Furthermore, if sulphides are present in these ores,

they can be first floated by the customary reagents without affecting the subsequent flotation of the tungsen. .When the tungsten ores are free of calcite and apatite, good concentrates are recovered by merely using an alkaline depressant for the siliceous gangue. When they are free of calcite, but contain apatite, the apatite may be depressed by an alkaline reagent in the rougher and ,an acid in thecleaner, or an acid reagent may be used in both the rougher and cleaner. The combination of a metal salt and soluble silicate in the rougher, and an alkaline or acid' Reference has been made above to the fact that heretofore the flotation of certain ores from calcite has been considered impossible. Specifically, the prior art is replete with statements to the eifect that fluorspar and barite as well as scheelite .could not be successfully separated from ores containing calcite. I

It is true thaton certain fluorspar ores, for instance, which contain small amounts of calcite, it has been possible to reject some of the calcite and produce fluorspar concentrates relatively free of this constituent. Usually, what success has been attained along this line has been due to the fact that the calcite was locked with the siliceous material and remained in the tailings ormiddling products. Furthermore, repeated cleaning of the rougher concentrates from an ore high in fluorspar but low in calcite does allow some of the calcite to drop back into the middling product as a result of a starving out process. This simply means that the abundance of fluorspar crowds the calcite .out of the froth. Even in such situations, however, much of the fluorspar remains in the middling products and consequently a low recovery is obtained.

When flourspar ores contain an appreciable quantity of calcite there has heretofore been no method by which concentrates low in calcite could be obtained. We are well aware that a certain flotation process is claimed to effect a successful separation of fluorspar from calcite. In this process the slime is removed either by a preliminary de-sliming operation prior to flotation, or. by frothing the pulp to remove the slime prior to floating the fluorspar. The sands are then floated in a warming pulp using cresylic acid and oleic acid and the rougher concentrates are cleaned four times. In this cleaning the calcite drops back in the middlings which are removed from the flotation circuit because their recirculation prevents high grade fluorspar concentrates from being produced. It is to be noted that in this prior process no actual depressant for calcite is used.

We have experimented with a large variety of fluorspar ores containing siliceous and calcareous gangue and numerous reagents have been tested. We have found several reagent combinations which will eflect a successful separation.

For instance, we have found that sodium silicate and an acidic reagent such as. acid dichromate are quite effective in depressing calcite. Acids other than the dichromate may be used in the flotation of these minerals. For instance, we have found that hydrochloric, nitric, sulphuric and formic acids specifically give equally good results. In using this combination on a fluorspar calcite ore assaying 73.60% fluorspar, 14.50%' calcite and 12.25% quartz, which was ground to 65 mesh, and floated in a mechanically agitated flotation machine, a concentrate was ootained assaying 97.71% fluorspar, 1.38% calcite, and 0.89% quartz. This concentrate accounted for 82.6% of the total fluorspar content of the ore. The reagents used were as follows:

Pounds per ton of crude ore Reagents Cleaners Rougher Oleic acid .96 Sodium silicate 40. 4' .0 1.00 Acid dichromate l. 50 2.00

calcite and 11.02% quartz.

' cate was used in the rougher.

spar.

We found that a silicate followed by dichromate effectively depressed the calcite and siliceous materials, providing a minimum amount of collector is used in the rougher. Too much oleic acid, or other fatty acid or derivative, produces rougher concentrates which are impossible to clean. It will be observed in the above that this process produces fluorspar concentrates from ores containing considerable calcite Without the necessity of a preliminary desliming or frothing operation, and that the calcite is actually depressed rather than merely starved out.

We have found that the order in which the reagents are introduced is immaterial, and, for instance, dichromate may be used in the rougher and silicate in the cleaners. As an-example of such a process, we secured a concentrate assaying 98.18% CaF-z, 1.06% CaCOa, and 0.56% SlOz, which accounted for 71.6% of the total fluorspar. The crude ore contained 69.93% fluorspar. 18.02%

The tailings contained less fluorspar than in the test'where sili- The rougher concentrate was cleaned first with dichromate and. then with silicate. The calcite was depressed in the middling and high grade fluorspar concentrates were produced as above indicated. In this test the ore was ground to 65 mesh and a mechanically agitated flotation machine was used. The groundore was first conditioned with dichromate before adding the oleic acid in the rougher. The amounts of reagents used were as follows:

sodium silicate, oleic acid and acid dichromate. In this process the pulp was conditioned with copper sulphate for several minutes and then the silicate was added. After a thorough mixing the oleic acid was introduced for floating the fluor- The rougher concentrates were first cleaned with silicate and then with dichromate. The amounts of reagents employed were as follows:

Pounds per ton of crude ore Reagents C d Cleaners on 1- tioner Rough Copper sulphate 1.00 Sodium silicate 0. 40 0. 40 Olelc acid Acid dichromate.

In using this reagent combination on an ore assaying 63.78% CaFa, 15.73% CaCOa, and

' 19.96% S102 and which had been ground to 65 mesh, a concentrate assaying 98.25% CaFz, 0.86% 09.00: and 0.78% $102 was recovered, which accounted for 85.6% of the total fluorspar in the ore.

We have found that when the proper amount of oleic acid and the metal salt-silicate combina-. tion is used, the silicate alone effectively depresses the calcite to give a high grade fluorspar concentrate. The exact amount of reagent for best results depends on the character of. the ore and must be determined by experimentation. Generally speaking, however, ores which contain considerable calcite or which require flne grinding for liberation require more metal salt and silicate for a separation.

While not wishing to be limited by the theory of the function of the metal salt-silicate combination, we believe that the metal salt, used in conditioning, reacts with the carbonate mineral to form a metal carbonate coating, and that the addition of silicate superimposes a coating of the metal silicate on the carbonate particle. Such coating may or may not be continuous, depending on the amount of reagent used and the time allowed for the reaction to proceed. The more complete the coating, the less tendency the particle has to float, and, in any event the silicate coated carbonate particle acts in flotation as a true silicate mineral. Inasmuch as the fluorspar does not take on the silicate coating, it is floated by the fatty acid and the same thing is true with respect to the scheelite above discussed and the barite which will be discussed later. If the calcite is insufilciently coated with the silicate, it will tend to float with the rougher concentrate, and in such event a cleaning with additional silicate depresses such calcite or siliceous material that may have floated either for this reason or because an excess amount of collector was used. A final cleaning with acid dichromate effectively depresses the remaining calcite and siliceous minerals.

The metal salt-silicate combination is effective on ores of any grade and the ores high in calcite are as amenable to treatment according to our invention as ores containing only small amounts of calcite. As an example, we conducted tests on a tailings product from a mill in the Kentucky- Illinois fiuorspar district which assayed 17.28% fluorspar, 73.61% calcite and 8.05% quartz. The clean mineral, of course, had been removed without dlfliculty in a gravity mill. Inasmuch as'flne grinding was necessary to liberate the fluorspar from the gangue which was composed of siliceous and calcareous minerals, concentration of such liberated fluorspar by gravity methods was impossible. I

This Kentucky-Illinois ore contained some sulphides which were first removed .by the conventional sulphide flotation reagents, sodium, aerofloat and aerofloat No. 25, and had no harmful effect on the subsequent fluorspar flotation. A fluorspar concentrate assaying 96.60% CaFz, 2.84% CaCOa and .01% SiOz was obtained which represented 48.6% of the total fluorspar. The flotation tailings assayed 7.57% CaF2, which represented 19.1% of the total fluorspar in the ore.

The reagents employed, and the amounts, were i as follows:

Bounds per ton of crude ore Reagents I Cleaners Sulphide rougher Rougher Sodium aerofloat Aerofloat No. 25.- Copper sulphate- Bodium silicate. Oleic acid Acid dichromate The pulp .vas conditioned with silicate before the oleic acid was added. As is apparent from the foregoing, good concentrates were obtained relatively free of sulphides, calcite and siliceous materials.

Barite Our experiments have shown that barite can also be separated from calcite, in much the same way as fluorspar, the metal salt-silicate combination successfuly retarding the calcite without affecting the barlte.

In a typical test, the ore was ground to pass mesh for liberation of the mineral and gangue, and the pulp was first conditioned with copper sulphate. The reagents used were as The concentrate recovered assayed 96.98% barite and 0.76% calcite, and accounted for 82.06% of the total barite in the feed. The bulk, 66.7%, of the calcite was depressed in the rougher.

' Fluorspar-barite ores During the course of our experiments, we found that the barite was sensitive to dichromate, and that slight excesses of this reagent depressed the barlte. Such was not the case with fluorspar, which was hardly affected by even large amounts of acid dichromate in the cleaners. This indicated that there was a possibility of separating fiuorspar from ores containing barite and calcite, and in order to verify this, we conducted a series of experiments.

We were able to effect such a separation, as ,is evidenced by a testin which a fluorspar-baritecalcite ore, assaying 41.82% fluorspar, 19.19% barite, and 33.80% calcite, gave a concentrate of 82.56% fluorspar, 13.24% barite and 1.49% calcite, andwhich accounted for 84.1% of the total fiuorspar and 29.4% of .the total barlte. The first middling, constituting 10% of the heads, assayed 23.56% fluorspar, 34.48% barite and 37.52% calcite; and the second and third middlings, aggregating 10.4% of the heads, assayed 11.21% fluorspar, 73.01% barite and 9.86% calcite. The reagents used were as follows:

Pounds per ton of crude ore Mineral carbonates generally As stated, it has heretofore been considered impossible to separate metal carbonates from calcite. Although the carbonates react differently in soap flotation, a mixture of free carbonates appears to behave in a manner that is intermediate between the reactions of its individual constituents, and. these constituents all float simultaneously rather than in succession. Our early experiments confirm the belief of the prior art.

that rhodochrosite and the manganese dioxides may be separated from calcite by using the above described metal salt-silicate combination, and a careful control of the oleic acid. As a matter of fact, we. believe that a successful separation of metal carbonates from calcite or the alkaline earth carbonates generally is possible, and we have no doubt that fluorspar, barite and scheelite can be successfully floated from carbonates other than calcite. I

It will be appreciated from all of the foregoing that we have devised a froth flotation process which is adapted to recover valuable constituents of ores that have heretofore been considered non-amenable to satisfactory flotation. Our invention is economical and easy to carry out. The concentrates recovered are of a high Recent tests made by us, however, indicategrade, and the materials are effectively depressed.

The processes herein set forth should be applicable to large bodies of ore that up, to the present time have been considered unrecoverable, and also to large quantities of tailings.

It will be apparent to any person skilled in the art that our invention is susceptible of a. wide range of equivalencies, and that various modifications may be made in either the steps, the reagents, or the quantity of reagents, without departing from the spirit of our inventive concept.

-In certain instances we have specifically listed equivalent reagents, but it will be apparent in connection with practically every reagent, our

range of equivalencies is not to be limited to such expressed instances. For instance, where we have used the term fatty acid in the specification or claims, it will 'be appreciated that this will include not only fatty acids per se but also fatty acid derivatives, such as soaps, or both.

We claim:

1. A method of floating minerals from calcite comprising introducing into the mineral contain- 7 ing calcite, a collector having 'a fatty acid radical,

morphic scheelite ores comprising introducing into the mineral containing calcite, a collector having a fatty acid radical, and a depressing agent for calcite consisting of a heavy metal salt which forms a relatively insoluble carbonate and an alkaline silicate, agitating and aerating the mixture, and then removing a concentrate relatively free from calcite.

3. A method of floating minerals from calcite comprising introducing into the mineral containing calcite, a collector having a fatty acid radical and a depressing agent for calcite consisting of an inorganic acid and an alkaline silicate, agitating and aerating the mixture, and then removing a concentrate relatively free from calcite.

4. A method of concentrating fiuorspar that also contains calcite comprising introducing into the mineral containing calcite, a collector havlnfl a fatty acid radical and a depressing agent for the gangue and calcite which consists of a soluble silicate and acid dichromate, agitating and aerating the mixture, and then removing a concentrate relatively free from calcite.

5. A method of concentrating fluorspar ore containing calcareous and siliceous gangue comprising introducing into the mineral containing calcite acid dichromate as a conditioning agent, then floating the fluorspar ore by a fatty acid radical, depressing the gangue by a soluble silicate, agitating and aerating the mixture and then removing a concentrate relatively free from calcite.

6. A method of concentrating a fluorspar ore containing calcareous materials comprising introducing into the mineral containing calcite a metallic salt conditioning agent, a collector having a fatty acid radical and a depressing agent consisting of acid dichromate and a soluble silicate, agitating and aerating the mixture, and then removing a concentrate relatively free from calcite.

7. A method of floating barite ores containing calcite comprising introducing into the mineral containing calcite a metal sulfate, a soluble silicate, a collector having a fatty acid radical and a depressing agent consisting of acid dichromate, agitating and aerating the mixture, and then removing a concentrate relatively free from calcite.

8. A method of concentrating contact metamorphic scheelite ores comprising introducing into the mineral containing calcite, a conditioning agent of copper sulfate, then adding a soluble silicate to depress the calcite, introducing a collector having a fatty acid radical, agitating and aerating the mixture and then removing the concentrate relatively free from calcite.

9. A method of concentrating contact metamorphic scheelite ores comprising introducing into the mineral containing calcite a heavy metal salt which forms a relatively insoluble carbonate, then adding a soluble silicate and alkaline reagent to depress the calcite, introducing a collector having a fatty acid radical, agitating and aerating the mixture, and then removing a concentrate relatively free from calcite.

10. A method of concentrating contact metamorphic scheelite ores comprising introducing into the mineral containing calcite a heavy metal salt which forms a relatively insoluble carbonate,- then adding a soluble silicate to depress the calcite, introducing a collector having a fatty acid radical, agitating and aerating the mixture, and subsequently cleaning the concentrate thus obtained by a soluble silicate and an alkaline reagent. a

JULIUS BRUCE CLEMMER. ROBERT GIBSON OMEARA.