US2469418A - Producing silicon - Google Patents

Description

Patented May 10, 1949 PROPUCIN G SILICON Marcus M. Striplin, Jr., Florence, Ala., assignor to Tennessee Valley Authority, a corporation of the United States of America No Drawing. Application June 19, 194B, Serial No. 677,841

9 Claims. (Cl. 23-2235) (Granted under the act of March 3, 1883, as amended April 30, 1928; 370 0. G. 757) The invention herein described may be manulectured and used by or for the Government for governmental purposes without the payment to me of any royalty thereon.

This invention relates in general to the recovery of valuable materials for alloys and in particular to the recovery of substantially pure silicon from alloys in which the silicon has crystallized.

Elemental silicon, which is usually termed silicon metal, is produced industrially in electric arc furnaces by reduction of pure quartz or silica sand with pure carbon, such as wood charcoal or petroleum coke. This process is attended with high loss of silicon by volatllization, which results in high consumption of raw materials and electric energy and which concentrates the nonvolatile impurities from the raw materials in the product so that it usually contains suificient iron, calcium, aluminum, titanium, and other impurities to decrease its silicon content to about 95 or 96 percent. To obtain silicon metal of sumciently high purity for many purposes, it is customary to grind the electric-furnace product and to dissolve the impurities from the finely divided material with acids.

The principal object of the present invention is to provide a process whereby relatively pure silicon metal may be readily recovered from siliconcontaining alloys which can be produced by reduction of mineral substances at a relatively low cost as compared to the cost of producing silicon by reduction of silica. Another object of this invention is to furnish a method for the production of substantially pure silicon metal from lowgrade raw materials. A further object of this invention is to provide a process whereby raw materials containing silica may be used in the production of alloys substantially free of crystallized silicon. Other objects of this invention include the provisionof a method for the separation of alloys into component materials having greater value than the original alloy.

It is generally known that silicon and other materials crystallize from certain molten electricfurnace products, such as high-silicon ferrosilicon and aluminum-silicon alloys, when such alloys are cooled. If, for example, molten aluminum-silicon alloy which contains about equal quantitiesof aluminum and silicon is cooled, silicon begins to crystallize at about 1100 C., thereby enriching the liquid in aluminum. With further cooling to the eutectic point, which occurs at about 575 0., liquid that contains only about 12 per cent silicon in solution results. It has been proposed to make the eutectic alloy by first reducing siliceous aluminum ores with carbon in an electric furnace to obtain crude aluminum-silicon alloy, then cooling the crude alloy until liquid of the eutectic composition is obtained, and finally settling, filtering, or centrifuging the resulting suspension to separate the liquid from the solidified material. However, with such processes, the iron, titanium, and certain other impurities present in the rawmaterials employed in production of the crude alloy also crystallize with the silicon, thereby rendering it so impure that it cannot be used for the purposes to which silicon metal is usually put, I

In the process disclosed herein, relatively pure silicon metal and other valuable products can be made from certain impure alloys of silicon and various metals by cooling the alloy until it is completely solid, grinding the cold alloy in such a manner as to liberate the crystallized silicon from the other components of the alloy, and subjecting the ground alloy to a particular combination of physical operations designed to classify the particles according to size, density, and other characteristics.

In the practice of the invention, an alloy containing silicon in excess of the proportion required for the eutectic composition is first produced by any suitable means, the alloy is cooled to bring about crystallization of the silicon and solidification of the other components, the solidified alloy is crushed and screened to separate a relatively fine fraction from a relatively coarse fraction, and the fine fraction is subjected to a separation method based on the density of the material.

For example, an alloy of iron and silicon may be prepared by reducing a charge oi. iron and silica in an electric furnace. The proportion of silicon in the allloy must be higher than 61.5 per cent, which is the eutectic composition, to ensure the presence of free, crystalline silicon. The process may be applied especially to recovery of silicon from the fines ordinarily produced in commercial ferrosilicon operations.

As another example, an alloy of silicon and aluminum may be produced by reducing a charge of clay with coke in an electric furnace. Normally such an alloy will contain about per cent aluminum, 40 per cent silicon, 5 per cent iron, and 5 per cent titanium, which gives a considerable excess of silicon over that required for the eutectic composition. The silicon content can be increased by adding silica to the charge, but it is undesirable to go very far in this direction since difficulties such as loss of silicon by volatilization are encountered. The iron and titanium contents of the alloy may be further reduced, when desirable, by the use of very pure clays or other minerals. However, in certain cases, it may be preferable to produce an alloy containing titanium since, as will be shown, the titanium may be concentrated in a fraction which is of particular value because of the presence of this element.

The principal advantage in producing such an alloy as a step in producing silicon is that volatilization of silicon compounds is minimized by the presence of aluminum or certain other metals. Also, as is well known, the direct reduction y of aluminum from its pure oxide by the action of carbon is not feasible because of the loss of aluminum by volatilization and the formation of carbides, whereas the presence of silicon minimizes both of these effects. It is therefore possible to produce an alloy of these two elements with much less difiiculty and expenditure of electrical energy than would be required for the production of either element alone.

The alloy thus produced is tapped from the furnace and collected in any convenient manner so that it will solidify and cool to ordinary temperatures. The cold alloy is then ground by crushing until, for an alloy such as that described above, about half of the ground material will pass a standard Tyler sieve having 28 meshes to the inch. Crushing is employed rather than impact methods of grinding in order to avoid embedding the brittle silicon particles in the malleable eutectic alloy. Optimum values for both the sieve size and the amount of material passing the sieve will vary with the composition of the alloy, but for the alloy used as an example, the above values have been found to be satisfactory. The coarser fraction will be found to be riched in aluminum and poorer in silicon, iron, and titanium than the original alloy.

The portion of the alloy passing the 28-mesh sieve is then subjected to a separation method based on density. Heavy liquid media, centrifuging, or other common methods may be used. This step is ordinarily carried out in such a manner as to obtain a separation between particles having a. density greater than 2.4 and those having a. density less than this value, although the optimum value for this density will vary with the composition of the alloy. The lighter fraction from this separation will consist essentially of the free silicon content of the cold alloy and will be found to be of a higher degree of purity than the usual commercial silicon metal produced by electrothermal reduction of silica.

The impurities in the alloy, particularly the iron and titanium, will be found to be concentrated in the heavy fraction from this separation. It has frequently been found advantageous to subject this heavy fraction to a further separation on a size basis, whereby a further concentration of these impurities in the finer fraction is obtained.

In a typical operation of the process of the invention an alloy of the following composition was treated:

Percent Silicon 39.8 Aluminum 49.2 Iron 2.9 Titanium 4.2 Other 3.9

Total 100.0

The alloy was pulverized in a jaw crusher and the pulverized alloy screened on a 28-mesh sieve. The portion which passed'through the sieve was treated with a mixture of tetrabromoethane and nitrobenzene having a specific gravity of 2.43 to effect a separation between the lighter portion floating and the heavier portion sinking in this liquid. The heavier portion"was then further separatelfby screening one. ZOO-mesh sieve. This series of operations therefore resulted in four fractions. The weights of these fractions obtained from 1 pound of alloy and the chemical analyses of the fractions were as follows:

Composition, percent Size of fraction, Specific Weight of mesh gravity fraction,

Si Al Fe Ti 2. 43 0. 496 20. 9 75. 2 l. 8 2. 4 2. 43 0. 233 96. 7 2. 4 0. 2 0. 1 2. 43 0.055 43. 8 2i. 5 16. 8 19.1 2. 43 0 216. 38. 5 47. 2 5. 3 l0. 1

The first fraction shown above, because of its high aluminum content, is more suitable than the original alloy for use in production of aluminum by selective alloying or electrolytic methods, and because of its reduced silicon content is more suitable for use in deoxidation of steel, particularly for steel having a low silicon specification. The second fraction meets the specifications for commercial-grade silicon metal, and as such is considerably more valuable than the original alloy, especially because of its fine particle size. The third fraction has a sufficiently high titanium content to make it comparable to commercial ferrotitanium, which commands a relatively high market price. Although this fraction represents only about 5 per cent of the original sample, it

= ium, this fraction is very similar to the original alloy and may be used for any purpose for which the original alloy is suited, or it may be returned to the reduction furnace to be recycled through the process.

By the practice of the process of the invention it is thus possible to obtain other valuable fractions in addition to silicon of purity comparable or superior to commercial silicon metal. The value of the aluminum and titanium is increased by the concentration obtained in the process until the total value of the fractions other than the high-silicon fraction equals or exceeds the total value of the original alloy. Therefore, th only costs chargeable to the silicon are those involved in grinding and beneficiation. These are low as compared to the cost of production by the usual electric furnace method.

Another unexpected advantage obtained with respect to the silicon metal produced was that the calcium content was lower than is obtainable in silicon metal produced commercially from silica. This element is undesirable in silicon used for alloying purposes, but electric-furnace silicon metal contains amounts varying from 0.2 to 0.7 per cent Ca, and lower percentages are obtainable only by acid treatment of the electric-furnace product.

By the process of the invention, on the other hand, silicon fractions were obtained which contained less than 0.1 per cent Ca without any additional treatment.

The process of the invention therefore results in the production of silicon metal and other valuable products in an efilcient and economical manner. The success of the invention is due to the peculiar qualities of alloys containing crystalline silicon in their response to beneficiation methods. It would not be expected from consideration of the prior art that the degree of separation obtained would result from grinding the cold alloy and then subjecting it to beneficiation methods. In fact, previous investigators in working with alloys which contained no crystalline silicon have specifled that the alloy be heated before grinding and ground hot in order to make the constituent of lowest melting point brittle and easy to grind to a particle size smaller than that of the other constituents. This method applied to alloys containing crystalline silicon would defeat the purpose of the grinding; the silicon, which is in general the highest melting-point constituent, is also the most brittle, so that any method which would render the lower melting constituents more brittle would reduce the degree of separation obtained on screening the ground alloy.

In addition to the above, the success of the invention depends on the particular combination and sequence of beneficiation steps used. If separation on a size basis were the only method used it would not be possible to obtain either the highsilicon or the high-titanium fractions and if density separation were the sole procedure employed, the high-aluminum fraction would be eliminated. The combination of the two methods is therefore necessary. The particular sequence of steps is also important because the first separation on a particle-size basis removes a considerable portion of the alloy and it is necessary to treat only the remaining portion by the relatively more expensive methods of density separation. A different sequence of steps would thus be more expensive and less effective.

While the properties of these alloys have long been known, none of the methods which have been proposed for utilizing these properties in the separation of the alloy constituents has been as simple or effective as the method of this invention. Thus, the method, already referred to, of melting an aluminum-silicon alloy, cooling it until silicon crystallizes out, and then removing the crystals by filtering or centrifuging involves the difficulties of conducting a separation operation at high temperatures, and yields only the aluminum-rich fraction as a valuable product, the silicon-rich fraction being so impure as to be of no particular value.

A further modification of the process would be to separate the high-silicon fraction on a size basis at about 200 mesh. A test has shown that the minus 200-mesh fraction obtained by this procedure contains 98.0 per cent silicon, 1.0 per cent aluminum, 0.3 per cent iron, and 0.1 per cent titanium. The value of this fraction is enhanced by the higher silicon content and the smaller particle size. In fact, this fraction is equal in quality to silicon metal that has been refined with acids.

It will be seen, therefore, that this invention actually may be carried out by the use of various modifications and changes without departing from its spirit and scope.

I claim:

1. A process for the production of elemental silicon in high purity which comprises preparing a hot, molten alloy containing about 50 per cent aluminum and about 40 per cent silicon in admix- 6 ture with lesser proportions of iron and titanium: cooling said alloy until it is completely solid: crushing the cold alloyto such degree that about half the resulting particles are small enough to pass through a screen having 28 meshes to the inch; screening said particles and thereby separating them into a relatively coarse fraction substantially free from crystalline silicon and a relatively fine fraction rich in crystalline silicon; depositing the said fine fraction in a body of liquid which is non-reactive therewith and has a specific gravity of about 2.43; and removing floating elemental silicon of high purity from the surface of said body of liquid.

2. A process for the production of elementalsilicon in high purity which comprises preparing a hot, molten alloy containing about 50 per cent aluminum, about 40 per cent silicon in admixture with about 5 per cent iron and about 5 per cent titanium; cooling said alloy until it is completely solid; crushing the cold alloy to such degree that about half the resulting particles are small enough to pass through a screen having 28 meshes to the inch; screening said particles and thereby separating them into a relatively coarse fraction substantially free from crystalline silicon and a relatively fine fraction rich in crystalline silicon; depositing the said fine fraction in a body of liquid which is non-reactive therewith and has a specific gravity of about 2.43; and removing floating elemental'silicon of high purity from the surface of said body of liquid.

3. A process for the production of elemental silicon in high purity which comprises preparing a hot, molten alloy containingabout 50 per cent aluminum and about 40 per cent silicon in admixture with lesser proportions of iron and titanium; cooling said alloy until it is completely solid;- crushing the cold alloy to such degree that about half the resulting particles are small enough to pass through a screen having 28 meshes to the inch; screening said particles and thereby separating them into a relatively coarse fraction substantially free from crystalline silicon and a relatively fine fraction rich in crystalline silicon; depositing the said fine fraction in a body of liquid which is non-reactive therewith and has a specific gravity of about 2.43; and removing floating elemental silicon containing less than 0.1 per cent calcium from the surface of said body of liquid.

4. In a process for the preparation of elemental silicon in high purity wherein crystalline silicon embedded in an alloy consisting essentially of aluminum and silicon in eutectic proportions is separated from said alloy, that improvement which comprises crushing the crystalline silicon and alloy, cold, to a coarse powder, screening the coarse powder thereby separating it into a relatively coarse fraction substantially free from crystalline silicon and a relatively fine fraction rich in crystalline silicon, and separating silicon in high purity from said fine fraction.

5. In a process for the preparation of elemental silicon in high purity wherein crystalline silicon embedded in an alloy consisting essentially of aluminum and silicon in eutectic proportions is separated from said alloy, that improvement which comprises crushing the crystalline silicon and alloy, cold, to a coarse powder, screening the coarse powder thereby separating it into a relatively coarse fraction substantially free from crystalline silicon and a relatively fine fraction rich in crystalline silicon, depositing said fine fraction in a liquid having a specific gravity of about 2.4 and which is non-reactive with said fraction, and removing floating elemental silicon of high purity from the surface of said liquid.

6. In a process for the preparation of elemental silicon in high purity wherein crystalline silicon embedded in an alloy consisting essentially of aluminum and silicon in eutectic proportions. is separated from said alloy, that improvement which comprises crushing the crystalline silicon and alloy, cold, to a coarse powder, screening the coarse powder thereby separating it into a relatively coarse fraction substantially free from crystalline silicon and a relatively fine fraction rich in crystalline silicon, depositing said fine fraction in a liquid having a specific gravity of about 2.4 and which is non-reactive with said fraction, removing material floating on the surface of said liquid, subsequently screening the removed material on a screen having 200 meshes to the inch and thereby separating it into a relatively coarse fraction and a relatively fine fraction containing 98 per cent elemental silicon.

7. A process for the preparation of highly pure elemental silicon from clay which comprises directly reducing clay with coke in an electrically heated zone, withdrawing a resulting hot molten alloy comprising about 50 per cent aluminum and about 40 per cent silicon in admixture with lesser proportions of iron and titanium, cooling the withdrawn alloy to ordinary temperatures, crushing the alloy, cold, to a coarse powder, screening said coarse powder and thereby separating it into a relatively coarse fraction substantially free from crystalline silicon and a relatively fine fraction rich in crystalline silicon, depositing the fine fraction in a liquidwhich is non-reactive therewith and which has a specific gravity of about 2.4, and removing floating elemental silicon of high purity from the surface of said liquid.

8. A process for the preparation of highly pure elemental silicon from clay which comprises directly reducing clay with coke in an electrically heated zone, withdrawing a resulting hot molten alloy comprising about 50 per cent aluminum, about 4'0 per cent silicon, about 5 per cent iron and about 5 per cent titanium, cooling the withdrawn alloy to ordinary temperatures, crushing the alloy, cold, to a coarse powder, screening said coarse powder and thereby separating it into a relatively coarse fraction substantially free from crystalline silicon and a relatively fine fraction rich in crystalline silicon, depositing the fine fraction in a liquid which is non-reactive therewith and which has a specific gravity of about 2.43, and removing floating elemental silicon of high purity from the surface of said liquid.

9. A process for the preparation of highly pure elemental silicon from clay which comprises directly reducing clay with coke in an electrically heated zone; withdrawing a resulting hot molten alloy comprising about 50 per cent aluminum, about 40 per cent silicon, about 5 per cent iron and about 5 per cent titanium from said heated zone; cooling the withdrawn alloy to ordinary temperatures; crushing the alloy, cold, to'such degree that about half the resulting particles will pass through a screen having 28 meshes to the inch; separating said particles into a relatively coarse fraction substantially free from crystalline silicon and a relatively fine fraction rich in crystal- .line silicon by screening the same on a screen having about 28 meshes to the inch; depositing the fine fraction in a liquid which is non-reactive therewith and which has a specific gravity of about 2.4; withdrawing elemental silicon of high purity floating on the surface of said liquid; withdrawing that part of said fine fraction which has sunk in said liquid; and separating said last mentioned part into a relatively coarse fraction and a relatively fine fraction rich in titanium.

. MARCUS M. S'IRIPLIN, JR.

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

Mellor: Inorganic and Theoretical Chemistry, vol. 6, pp. 184, 185, 201, and 209. Published by Longmans, Green, 8: Co., London (1925).

Certificate of Correction Patent No. 2,469,418. May 10, 1949. MARCUS M. STRIPLIN, JR. It is hereby certified that errors appear in the printed specification of the above numbered patent requiring correction as follows:

Column 1, line 6, for the words materials for read materials from; column 2, line 40, for allloy read alloy; column 3, line 39, for riched read richer; column 4, line 17, in the heading to the table, third column thereof, after fraction, insert 1b.; same table and column, last line, for O 216. read 0.216

and that the said Letters Patent should be read with these corrections therein that the same may conform to the record of the case in the Patent Office.

Signed and sealed this 11th day of October, A. D. 1949.

THOMAS F. MURPHY,

Assistant Commissioner of Patents.