US2421648A - Bauxite beneficiation by hydration of gangue - Google Patents

Description

f ores which are found in this country, in the Arkansas and Appalachian fields.

, ores. Because "bauxite Patented June 3, 1947 navxrrn BENaFlcrA'rIon BY or GANGUE l v, Roy Alderson Pickens,

This invention relates to beneciation of low grade bauxite ores and comprises a continuationin-part of my copending applicationA for U. S. Letsl Patent, Serial No. 415,889, filed October 21,

So called bauxite ores, as sources of aluminum material and aluminum chemicals are com-` mercially of great importance. The economic factors involved in the commercial production of these materials have lresulted in the fixing of definite minimum requirements for the ore to be used forvarious purposes. For example, in the manufacture of metallic aluminum, the ore is normally required to assay over 50% A1203 and less than 7% S102. In the manufacture of abrasives, the S102 content normally must not exceed 6.5% and the FezO: content mustbe less than 7%. commercially `a premium is usually paid for ores which exceed the minimum speciilcations.

In the past, deposits of native ore which would meet or better these requirements have been available. However, these deposits oi' naturallyoccurringfhigh-grade ores are being very -rapidly exhausted. As a consequence more attention is being directed to methods and means for utilizing the extensive deposits of'lower Since the operations in which these ores are used as raw materials are to a large extent dependent upon the use of ores having the speciiled gradesy for satisfactory operation, there is a demand for a suitable process of 'beneiiciating these is normally a low-priced ore, such a process in order to be .practical must bef also relatively simple and cheap to operate. TheV process of the present invention presents one solution of this problem.

In considering the process of the present invention, an examination of theore as it is mined is helpful. Actually, the designation "bauxite ore is usually a misnomer. The principal constituents are generally found to be gibbsite or cliachite (AlaOafSHaO) and kaolinite ally assayina from 0.5 to 6.0% T102. Quartz sand (S102) is occasionally `present to the extent of 0.1 to 2.0%.

grade t principallynrnnATIoN Stamford, Conn., assigner to American Cyanamid Company, NewYork, N. Y., a corporation of Maine Y Application February 13, 1943, Serial No. 475,801

Physically, the dit-ferent bauxite ores resemble each other quiteclosely. Usually there is a core of gibbsite. In size, these cores may range from very small particles, to good sized marbles or larger.` 'I'hese cores are neither perfectly solid nor perfectly smooth. Consequently, .a certain amount of gangue is always found to bepresent in the pits and voids of thesecores. A considerably larger portion of gangue adheres ilrmly and screening, a part of the sangue which has been.` designated above as matrix can be readily removed. It will be apparent that its removal will raise the residual alumina-silica ratio. However, with low grade ores, such as thosewith which the present invention is primarily iconcerned, this will not ordinarily bring thealuminasilica ratio up to that necessary for an ore which is to be used, for example, in the production of metallic aluminum.A

These washing and screening treatments have been used to some extentin foreign countries.

Their use in such cases is merely to make a slight `saving on the freight charges 0n export ore.

Some savings can be made in this way even in the case of ore which is already of sulciently high grade for manufacturing purposes.. These washing processes have a. useful purpose in this I field even though the beneciation is so slight as vto seldom warrant the expense of using them solely for the purpose of raising the grade.

That portion of the gangue which adheres to the surface of the core is much more stably held than is the matrix gangue and will withstand considerable rough handling without being released. The gangue which is held in the pits and voids is even more firmly attached: Because some gangue may be present in a bauxite suit- .able for the production of aluminum products,

ore from some deposits can be raised to the necessary grade simply lby using the washing and screening methods set forth above. However, as pointed out, the degree of beneflciation which can be obtained is so small that these ores are normally considered as part of the rapidly vanishing high-grade ores. The process of the 4present invention is concerned vonly with those ores in which it is necessary to remove substantially all of the matrix and at least a portion of the firmly adherent gansue.

in order to effect a. com-v ciiic gravities, prevents acceptable beneficiation by ordinary commercial methods centration.

Froth flotation methods suffer from the draw-"f back that most of the available reagents exhibit little selectivity, possibly due to the aluminum content of kaolinite. Such reagents as are suffciently selective to warrant potential use are,` expensive and extremely sensitive to the presence of slimes. slimed very fully, an operation which the similarities in specific gravities makes both difficult and expensive. Successful commercial |beneficiation of such ores thereforerequires a special procedure adapted for that purpose. As pointed out, in addition to being operable, such processes must also be simple to use and cheap in operation if they are to be suitable for commercial development.

In my 'c'opending application 406,533, filed August 12, 1941, I have shown that such low grade ore may be satisfactorily beneciated into a commercially acceptable grade by a process which comprises preliminarily crushing the ore, soaking the orefor a considerable period, ne crushing the soaked ore, and desliming the crushed ore by wet screening or hydraulic classification. It was pointed out that one of the es. sential factors was to transform the clay into a physical condition such that it can be separated and removed as slimes by any suitable sixing operation. A soaking period, dependent in length on the nature of the ore and the extent of the preliminary crushing, was employed for Athis transformation. While extending the soaking period used with a particular ore tended toincrease'the silica removed in the slimes, it also decreased the over-all recovery of aluminum so that an economic balance between the two factors had to be determined and maintained.

This process using an extended soaking period will produce satisfactory results, but it is economically inconvenient both because ofthe waiting period involved and the extensive equipment l required for storage. In the process of my previous application a material amount of stirring and crushing was necessary after the soaking period and before the desliming procedure in order to insure proper suspension and size grading,

I have `now found that if agitation of the ore is l carried out during the soaking period the necessary time interval may be shortened to a wholly unexpected degree. With suillciently violent agi tation I obtained in a few minutes results comparable to those obtained with the same ore only' after a period of twenty-four to forty-eight hours of soaking without agitation.

This is of great commercial advantage in that it ,saves both time and storage equipment. In addition, I have found that no subsequent stirring step was necessary as in 'the case of the soaking without agitation and that the material may be passed directly to the desliming apparatus without any fine crushing. This is also a considerable advantage because of the saving in power and equipment needed for the crushing steps.

of gravity con-*f As a result, the ore must .'be de- 4 The drawing shows a representation of a ow sheet of the process.

In general, the object of the present invention is accomplished by crushing the ore to a size convenient for handling, making it into a pulp with uwater, agitating the pulp to separate the gangue from the other constituents and desliming the resultant mixture.

In carrying out the processof the present inventlon, I have found that the preliminary crushing of the ore is relatively unimportant. The principal limitation is that the material be small enough to pass into the apparatus used to carry outl theagitation. The process is found to work well with ores having been crushed sufficiently to pass a 1/4 to 1/2 inch screen. However, grinding the material to a somewhat smaller size is 'some times beneficial, especially where a relatively large proportion of the gangue is held in the pits and voids of the core.

Grinding the ore to .the extent of producing an average particle size less than 14 mesh has been found to yield excessive amounts of gibbsite and/or bauxite lines. The mineral values represented by these f-lnes are subsequently lost in the desliming operation. Consequently, excessive grinding must be avoided. This in itself is a considerable advantage because it eliminates the necessity for grinding the ore to a specific small size as is required in many beneiiciating processes.v ,l

The present process has an additional advan-Iv tage in that it is not particularly critical to the amount of water used to dilute the crushed pulp.

Only so much as is rrequired for convenient4 hanl dling is necessary. In general, the more violent the agitation the less water is necessary. Thisnot only makes the process simpler from a practical point of view, but is of considerable importance agitation may be carried out in any conventional machine including aerators, mechanical stirrers,.

beaters,bal1 or rod mills, pumps and the like. However, these machines such as the mills which normally involve grinding or disintegration must 'be carefully used. Since as pointed out above, an.

excessive reduction in the average particle size increases the alumina loss in the wet slimes, in using a ball and rod mill for example, the balls and rods should be either entirely eliminated or only one or two inserted to provide a slight agita- A tion with a minimum of reduction in particle size.

Goods results are obtained by using a mill simply as a rolling drum.

Certain machines such as mechanical-type flotation machines which produce a very violent agitation were found to give equivalent results in a correspondingly shorter time than that required when using a more gentle agitation. Excellent results were obtained, for example, using a Fagregren flotation machine. Favorable ores gave acceptable results in as short a period as one minute. Most of the ores tested required no longer than from 1-5 minutes and only a very few particularly recalcitrant ores required as long as 30 minutes. Aside from the shortened time period permissible with a flotation machine, they possess a number of other specific advantages. These machines have a tremendous capacity ascompared to other types of stirrers. 'Ihe necessary amount of water can be, to a large extent, recovered and reused, which would be of advantage in locallties where water supply is a serious problem. Again, due to the exceedingly short time period required, no silicagel is formed and does not interfere in the handling of the waste. Nor is the time period required long enough to allow any appreciable hydration of the alumina Dresent, so that there is veryilittle .loss in theslimes due to this cause.

After the agitation. the pulp may be directly deslimed either by a wet screening operation or by hydraulic classification. If desired the deslimed sands may be subjected to aJ tabling operation which will further increase the aluminum ratio by removing a portion of the heavier minerals, such as those of iron or titanium. For abrasive manufacture, the bauxite must give a.A calcined product containing less than 6.5% silica and less than '7% iron as ferric oxide. My beneilciated bauxite is enable its being readily brought limits by such tabling steps. Y,

While it is not proposed to limit the invention to any particular theory of operation, it is believed that the actual concentration, i. e., the increase oi the alumina-silica ratio obtained by the process of the present invention is due to two factors. The rst of these is the separation of a suicient amount oi adhering gangue from the mineral by the action of self-attrition which occurs during the agitating step. Both bauxite and gibbsite are slightly harder than the silicatebearing gangue. arated from the minerals by the action ofthe particles one upon the other without materially decreasing the average-particle size of the mineral values themselves. So far as the externally adhering sangue is concerned, the resultant action is comparable to peeling an onion by removing successive external layers.

The second important factor is apparently the causing of the sangue to hydrate to a more readwithin these ily water disnersible form. Tins permits the exi0 cess adherent sangue to he more readily looscned, and separated from the cleaned mineral values by the action of 'self-attrition. As distinguisheol from my previously mentioned, cepending application in which the hydration of the gang-ue is of primary importance, in the presentV process it appears to loe secondary to and in aid of the action of self-attrition.

The invention will be described in greater de` tail in conjunction with the following specific examples which represent the treatment of typical low grade bauxite ores. These examples are merely illustrative and are not meant to limit the scope of the invention. The parts are by weight. l

EXAMPLE l In order to test the effect of preliminary crushing before subjecting the ore to the agitating and desliming operations, a number of samples of low-grade bauxite ore from Arkansas were stage-A crushed through several designated screen sizes. The crushed ore was diluted to approximately solids with water and subjected to a 5 minute treatment in a Fagregren flotation machine from which alternate bars of the rotor and'stator were removed. The air inlet was closed'.` The agitated materials were wet screened with a seperation at plus 65 mesh. The undersize portion was settled with sodium silicate as a dispersant low enough in silica to Therefore, the sangue is sept the size of the iced.

for one minute, decanted and the residue washed twice by decantatlon. The results are shown in Table I.

Table I Assays. Per Cent Per Loss Per Cent' Distribution Size Product (cht oitiig- Ahoi S102 A1303 S101 .525"- Feed 100.00 20.27 51.70 12.38 100.0010000 .525 me 44.55 29. 23 51.47 5.39 44.35 19.40 .525" Total Sand. 59.01 23.61 51.42 6.09 513.6912002 .020" Feed 100.00 158.01 52.46 12.56 10000100.00 .020". +65 mesh" 42.26 58.01. 53.57 5.37 43.16 41.93 .020" Total Send- 55.72 @.48 52.80 5.90 56.20 26.00

0mesn Feed 100.00 anos 52.00 12.05 `10o-00100.00 mesh... +65 mesh.. 44.64 28.56 52.92 6.09 45.37 21.00 6mesh Totalsand. 56.85 428.19 52.42 6.42 57.22 28.18 10 mesh Feed 100.00 28. 42 52. 12 12.27 100.00 100.00 10mesh |ES5mesh 44.39 29.25 52.95 5.48 45.10 19.83 10mesh.. Total sand 55.46 28.60 52.47 5.80 55.65 26. 67

14`mesh.- Feed 100.00 26.40 51.70 12.44 100.0010000 l4mesh.. +5mesh.- `42.08 29.13 52.27 5.73 43.75 19.67 lfimesh.- Total Sand- 57.23 28.52 51.35 0.24 57.44 28. 76

lit will be noted that in each case trom approximately 7l' to 74% 'oi the total silica was reiected and the weight recovery oi total sands was excellent at from 55 to 50%. These results indicate that preliminary crushing is important only in order that the larger elses in the aggregate can be reduced sunlciently to nass between the bars ofthe rotor and stator. llnilormly successiul silica rejection appears to have been obtained combined with good recovery regardless of limitante 2 ln order to show the eect of excessive grinding without the necessary degree lof agitation to produce effective hydration, a number of samples of the saine ore were carefully stage-crushed in a wet rod mill to nass a 20 mesh screen; 'the samples being ground in three stages with :the undersize being removed `between each stage. In samples l and 2, sodium hydroxide was used to maintain a basic medium and with the third sample, .Casi was used. The ground ore was diluted to about 15% solids and rieslimed by settling and decantation. The results are shown in Table lli.

Tome .U

Asso o, Per Cent Per Loss Per ant Distribution Sample Cont on llg- Wt. nition A1200 S102 A100: SiO:

Feed l. 100. 00 20. 00 5l. 20 12. 55 100.00 100.00 l Slime 55. 2B 20. 22 50. 86 16. 25 54. 91 7l. 56 Sand.. 44. 32 27. 44 5l. 02 7. 98 45. 09 2S. 44 2 Slime 62. 06 20. 93 51. 07 15. 31 02. 50 75L 44 San 37. 34 09 51. 42 7. 92 37. 50 23X. 56 3 Slime B2. 10 26. 50 51. 90 15. 33 02. 95 75. 84 Sand.- 37. 27. 78 50. 05 8. 00 37. 05 24. 15

.l Assay average oi many samples. Grinding: 3-5 minuto stages with -20 mesh removed after ilrst and second stages. Desliming: Settling and docantation. g

. EXAMPLE 3 A similar sample of the same ore was crushed to pass a 14 mesh screen. One portion of the minus 14 mesh ore was diluted to about 20% solids and wet screened without further treatment. A second portion was also diluted to 20% solids and subjected to a. 5 minute treatment in a Fagregren flotation machine with the air shut oi and then deslimed by wet screening. The results which are shown in the following Table III, clearly show the increased beneciation produced by the process of the present invention over that obtainable by crushing and washing alone.

metallic aluminum. In the shortest period the residual silica content was just within the commercial limitation, whereas in the two longer periods the residual silica content was correspondingly-lower. However, the longer periods Table IH 5 resulted in excessive loss of alumina andthe handling of the material was complicated by Time of Per Loss Ptsyeft Digglgilggm the formation of considerable silica gel.

Ato?. (150?? giltiloge Exam 5 41,0. sto, .41.05 e101 10 LE In order to show the saving in time obtainable Feed* 10o-00 2G-99 51420 12-55 10o-00 100-00 by the use of a iiotation machine, another por- 0 200 meglL 4273 24,86 50 l5 l9 78 41,33 67,33 tion 0f the Same Cl'uShed Ore 0f Example 4 WaS +200 mesh- 57-27 2&57 5h98 7,16 5&14 3167 diluted to approximately 20% solids in water min--. 2005.5011. 50.01 24.12 40.02 10.00 40.00 01.02 15 and subjected to a ve minute treatment in a +200 mh- 49-13 29-33 5253 485 50-40 1&93 Fagregren flotation machine with the air shut 01T. The material was then wet screene, cuts Assay mi many Sample* being made at plus 65 and plus 200 mesh. 'I'he It should be pointed out that the assays of both results are shown in Table V.

Table V Assays, Per Cent Distribution, Per Cent Time lProduct perwflft lsig AhQs F6101 SiO; A1101 F6203 lSiO,

o Feed 100.00 00 51.20 0.04 100.00 100.00 100.00 5 Min. m Flotation Mamme-.. +05 mesh-- 37.37 20.72 53.07 10.04 4. 5s 33. 73 54.00 13.04 +200 (01001)... 40.13 20.33 52.53 v10.30 4.35 50.40 72.05 10.00

vfeed and product are expressed in terms of total As will be seen, both the plus 65 and the total A1203 and total Si02. Since most of the silica plus 200 sands are commercially acceptable for is present as kaolinite, for each unit of Si02 disthe production of metallic aluminum and are parcarded there is a necessary loss of 0.85 unit of ticularly low in silica content. .The total plus 200 equivalent alumina. Consequently, the recovery 4o sands arealso commercially suitable from the of A1203, based on available A1200, is actually point of view of total recovery. No gel is formed much higher than the 50.40 shown in Table III. in this treatment. The shortened time of live l EXAMPLE 4 minutes practically eliminates the entire soaking period. The Fe203 content can readily be brought In order to show the effect of varying the period below 7% by tabllna without raising the Silica of agitation another sample of the low-grade ore above the '65% maxlmum .required for abrasive was crushed to about 14 mesh, diluted to about manufactura. 15% solids with water, and subjected to gentle agitation in a rolling drum without any mixing EXAMPLE 6 or grinding element such as rods or balls being In order' to illustrate the beneflclatlng possiused; The Crushed Ore was divided into Several 5 biiities 0f the agitation treatment, a sample 0f p0riln- Three POItOnS Were agita-ted OI'- a' low-grade` Georgia`bauxite ore was crushed to period of fifty minutes, 191/2 hours, and 48 hours pass a 1/2 inch screen, diluted with water to about respctvel In each CaSe the Pulp Was Wet 15%V solids and agitated in a slowly revolving screened, cuts being made at plus 65 and plus 200. 5. drum for 92 honra The pulp was than desllmed The results are Shown in Table IV- 5 by settling and decautation. Desliming was rendered diiiicult by the presence of considerable Table IV quantities of gel formed during the rolling treatment. The results are shown in the following Assays, Distribution, 60 table- Per Loss Per Cent Per Cent l Table VI Time Product Cen on Ig- WL nifinn A110, sto, A140. sio, l

' psiit D'tilin' 0 Feed 100.00 20.00 51.20 125510000111000 65 Pelnt {Lgstglll mm +05mes11- 42.23 20.10 52.00 017.43.40 20.77 .4140, S10, A140. sio.

+200(t0te1) 55.13 20.03 52.42 0.03 50.45 20.34 10041100150-- +05mes11- 30.75 20.00 51.75 4. 01 37.14 14.00 Feed 100.00. 20.40 55.12 14.08 100.00 100.00 +200 (50501).- 4s. 40 20.23 51.45 5.13 4a 72 10. s2 s11me 75.00 25.07 53.04 17.52 73.02 00.04 sena 24.04 30.57 50.01 5.72 20.03 0.30 40110urs +05mes11 30.11 30.10 51.70 4.44 30.51 12.77 7o +20o(t0ta1) 40.14 20.75 51.51 4.77 40.42 17.55

` EXAMPLE' 'I It will be noted from these results that in each In order to illustrate the effect of short-time case the total plus 200 mesh sands were of comagitation, a number of samples of the same low fmercially acceptable grade for the production of m grade Georgia bauxite ore were crushed to nassl a aanwas g/2 inch screen and made into a sludge at about 35% solids. These were then agitated for il, 3, 10 and 20 minutes respectively with a mechanical stirrer, stirring at a rate sufficient to produce a definite boiling action and adding any necessary water to insure ready agitation as the sludge thickened. Desliming was effected lby settling and decantation. The results are'shown in Table VII.

Table V1.1'

A Assays, Distribution, Per Loss Per Cent Per Cent Time Cent on Igi Wt. nition A1203 S105 A1203 S102 Feed-... 100. 00 26. 60 55.04 14. 07 100.00 100.00

Slime 10.42 20.23 47. 42 28. 34 8.98 20.13' Band.- 89. 58 27. 34 55. 93 13. 08 91. 02 79. 87

Bmn Feci--. 100.00 26.51 54.44 15.27 100.00 100.00 Slime- 21. 85 20. 55 47. 80 27. 69 19. 18 30. 63 Sand.. 78. 15 28. 18 56. 30 11. 79 80. 82 60. 37

min Feed- 100.00 26. 59 55.32 14.41 100.00 100.00 Slime- 35. 29 21.94 49. 46 24. 87 3l. 55 60. 89 Sand..` 64. 71 29.12 58. 52 8. 71 08. 45 39.11

2o mmm.- Feed-.. 10o. o0 ze. te et. 4e 14. 21 100.00 100.00'

Slime- 41. 42 22. 49 50. 19 23. 62 37. 49 08. 87 831111..-- 5B. 88 29. 97 59. 18 7. 55 62. 51 31. 13

EXAMPLE 8 in order to demonstrate that/the agitation die more than merely remove an increased amount of nes, a screen analysis of the sands obtained from the non-agitated sample was made. The

results are shown in Table Vm.

1. A method of beneficiar-.ing low grace bauxite ores, which ores comprise bauxite-minerals par- Atitles aggregated with an excess of rmlyattached hydratable silica-bearing sangue, to produce a high grade bauxite concentrate suitable for the production of metallic aluminum and assaying not more than 7% silica, which comprises: cmshing the ore to from about minus 1/2 inch to plus 14 mesh under such conditions as to produce a minimum of minus 14 bauxite mesh lines; forming the crushed ore particles into an aqueous pulp, whereby the hydratable sangue on the particles surfaces is wetted and transformed to a hydrated water-dispersible form; agitating the aqueous pulp under conditions producing essentially self-attrition only, whereby the hydrated gangue is detached and dispersed immediately hydration is complete without producing excessive size-reduction of the bauxite particles; continuing the agitation and self-attrition in the presence of water until a. major portion of the hydratable gangue becomes hydrated and dispersed; and deslimng the treated pulp in the presence of water.

2. A process according to claim l in which the desliming operation is carried out by hydraulic ciassliication.

3. A process according to claim l in which the desliming operation is carried out by washing and screening.

RDY ALDERSON PICKENS.

REFERENCES einen The following references are of record in the file of this patent:

UNITED STATES PATENTS Number Name Date 1,397,414 Everhart Nov. 15, 1921 FOREIGN PATENTS Number Country Date 424,514 France Mar. 15, 1911 OTHER REFERENCES Conditioning Surface for FlotatiomNorman 8: Ralston. A. I. M. E. Tech, Paper 1074.

Bureau of Mines Report of Investigation No. 290, Gorno Obogationo Delo, 1932 No. 11, pages 23-34. (Copy in Geological Library of Survey.)

Bureau of Mines Bulletin No. 312, pages 22, 23', and 68.

Images