US2076067A - Process of making beryllium - Google Patents

Description

Patented Apr. 6, 1937 UNITED STATES PATENT PROCESS OF MAKING BERYLLIUM poration of Ohio No Drawing.

5 Claims.

This invention relates to the making of beryllium metal, and an object thereof is the provision of a novel method of making beryllium in pure form, with greater ease and convenience, and greater economy of material, than has heretofore been possible.

Other objects and advantages of the invention will be apparent to those skilled in the art from the following description thereof.

.In general, the present process comprises treatby dissolving the salts and leaving behind flake or powdered beryllium. The metal is then compressed and melted under a suitable flux or in vacuo in accordance with the methods'now' well known to those skilled in the art. When salts of other metals are mixed with the beryllium salt they are preferably water soluble, and either incapable of being reduced by the reducing metal used or of such a nature that the end products of their reaction will not alloy with beryllium and'preferably be water soluble. The reducing metal is preferably so chosen that in the end reaction it produces a water soluble salt, in which event it is merely necessary to leach the product of the reduction step with water to obtain the beryllium metal where the salts of the other I .etals used are of the aforesaid preferred character.

Among the salts of berylliumthat are suitable for use in this process are the halides, and of these the chloride is preferred. The fluoride is least suitable because the product of its reaction with the most desirable of the metals that can be used to reduceit is such that the beryllium cannot be freed therefrom by leaching. The bromide and iodide are usable, but economically unsatisfactory because of their low metallic content and high cost.

. Since the beryllium halides are quite volatile at temperatures only a little above their melting points, it is not economical to melt them alone except in tightly closed vessels, hence it is preferable to dissolve them in other low melting point salts or mixtures of salts. As previously stated, these other salts should preferably be water soluble and either incapable of being acted upon by the metal chosen as a reducing agent, or of such a nature that the end products of their reaction will not alloy with the beryllium and preferably be water soluble. Many of the simpler salts of the alkali or alkaline earth metals, such as the chlo- .ture.

Application May 25, 1935, Serial No. 23,519

rides of sodium, potassium, and lithium answer this requirement. Since the beryllium halides, and especially beryllium chloride, are quite volatile even from fused solutions or 'mixtures at higher temperatures, it is desirable to use a fused salt mixture of relatively low melting point in order to prevent the vaporization of the beryllium salt, especially while it is being added to the mix- Lithium chloride, because of its low melting point, may be used alone as a diluent for the beryllium salt, but sodium and potassium chlorides are preferably used in their low melting point eutectic mixture.

Magnesium, in addition to being relatively cheap, abundant and available, is altogether the most suitable reducing agent. Its reaction with the beryllium salts, such as the chloride, dissolved in a fusion of other salts such as sodium chloride and potassium chloride is very sharp and. definite but comparatively mild and unexplosive. The reaction takes place over a large range of temperatures starting well below the melting point of magnesium. The products of the reaction are readily water soluble, permitting the easy separation of the beryllium formed by means of leaching with' water. The magnesium displaces the sodium and potassium is limited by the almost explosive violence of the reaction produced, for which reason they are not well adapted to large scale production. Also the beryllium produced with these metals as well as the alkaline earth metals is preponderantly very fine or fog and consequently oifers difl'iculties in the later melting steps.

In carrying out the preferred embodiment of the process of the present invention an anhydrous beryllium chloride is produced in any well known manner, such as the method described in my earlier Patent No. 1,805,567, or it may be produced by methods analogous to those employed in producing the very similar salts, aluminum chloride and magnesium chloride, which are now consumed in large quantities in industry. A fusion is then made of approximately equal parts of sodium chloride and potassium chloride, and this is brought to a temperature only slightly above its melting point. To this fusion the beryllium chloride is added. The ratio of beryllium chloride to the sodium and potassium chlorides is not critical. Fairly satisfactory operation results when the beryllium chloride is any here within the range of 5% to 75% by weight 0 the total fused salt mixture, but the best yields are obtained'with a beryllium chloride content of about 25%. The greater dilution of the beryllium chloride favors a more complete reaction with the reducing agent, the formation of larger and therefore more useful flakes of beryllium, and also causes less loss by vaporization of the beryllium chloride. After the beryllium chloride has been added, the resultant mixture has a very much lower melting point, which may be well below red heat, depending on the concentration of beryllium chloride.

The vessel in which the melt is made may be of any convenient material which does not react with the reducing metal as previously stated. Magnesium metal is the preferred reducing agent, and when this is used the pot or vessel may be of iron, or ferro-chrome. However, this introduces a minute amount of iron as an impurity in the beryllium, and if greater purity is required a vessel or pot made of porcelain or fused silica may be used.

After the above melt is completed, the magnesium metal is added, preferably in the form of thin chips, granules, ribbon, or the like. The reaction takes place at any temperature at which the salt mixture will remain liquid. For example, the eutectic mixture of sodium chloride and beryllium chloride melts at approximately 225 C., and the beryllium chloride reacts with magnesium even at this low temperature. However, the preferred temperature is about or slightly below the melting point of magnesium. The magnesium is added slowly and the salt bath is stirred constantly during the reaction. The reaction between the magnesium and the beryllium chloride takes place evenly and smoothly and without violence. As the magnesium chips enter the bath they are observed to glow and then disappear quickly, while the temperature of the bath rises slowly.

The size of the magnesium particles as added and the stirring of the melt during the reaction are both matters of considerable importance. If large pieces of magnesium are used, they become coated with a layer of beryllium which seems to hinder further reaction, and leaves a core of unattacked magnesium. Introducing the magnesi um in fairly small pieces and stirring constantly so as to prevent the beryllium from clinging to the magnesium and thereby protecting it, obviates this difiiculty and promotes a complete reaction. It is probably for this reason also that the reaction proceeds better at a temperature slighty below the melting point of magnesium since melted magnesium tends to flow together into masses which expose too little reaction surface.

The amount of magnesium added may be the theoretical calculated weight or slightly less, say 95% of theoretical, to allow for slight loss of beryllium chloride by vaporization during the melting and reaction. The magnesium seems to displace the beryllium quantitatively, and at the temperature employed it shows no tendency to alloy with the beryllium. The end point of the reaction may be observed accurately by the fact that after the reaction is complete any additional particles of magnesium added cease to glow. During the melting and reaction a slight draft may be maintained over the reaction vessel, and directed into a cooling chamber. This draws in and condenses the salts lost by vaporization and they can then be recovered and reworked. The entire reaction requires only a comparatively few minutes, and at the completion of the reaction the beryllium is present in the melt in the form of loosely coherent sponge. If care is used, this sponge stays at the bottom of the vessel and the clear supernatant fused salt may be decanted. The remaining small part of the fused mixture containing the beryllium metal is now permitted to cool, after which it is crushed and leached with water. The products of the reaction and the salts of the melt in this instance, are beryllium, magnesium chloride, sodium chloride, and potassium chloride. All of these with the exception of the beryllium are water soluble and can be readily separated from the beryllium-by leaching.

The filtering is done first through a screen of about mesh, then through a screen of about 325 mesh, then through filter cloth or filter paper. If any considerable excess of magnesium has been used it is usually present in globules or agglomerates and is retained on the 50 mesh screen. Part of the washing is done with hot water, and the finely divided excess magnesium, if any, is attacked by the hot water and may be decanted as hydroxide. In general, however, it is preferable to use a slight deficiency of magnesium, making these precautions unnecessary. The metallic flakes remaining on the 325 mesh screen are all useful for melting into massive ingots, and usually constitute about 75% or more of the theoretical metallic content of the beryllium chloride. The very fine metal that goes through the 325 mesh screen and is retained on the filter cloth or paper can also be melted into massive ingots, but only with considerable loss. It is therefore more economical sometimes to return it to the chlorinator and rework it into chloride.

The metal that remains on the 325 mesh screen is preferably compacted into suitable masses by means of high pressure and then melted either in vacuum or under a suitable flux. This results in a loss of about 10% most of which passes into the dross or flux and can be reworked. With these various simple recovery steps the yield of massive melted metal can be made to reach or exceed 95% of the theoretical metalcontent of the chloride. The metal so made is exceptionally pure, possesses considerable malleability, a property not commonly shown by beryllium made by any of the processes heretofore used, and not definitely heretofore known to exist, and is especially free from carbon or iron.

The decanted fused salt may be used as a diluent for more beryllium chloride, and another reaction may be carried out, and so on. During this process the salt becomes progressively richer in magnesium chloride, but this does not affect it adversely. At a suitable period in the cycle, this anhydrous fused salt mixture containing magnesium chloride may be poured into an electrolytic cell and electrolyzed to free the magnesium which may then be used over. The chlorine evolved may likewise be used over in the chlorinator. Through this cyclical recovery of the chlorine and magnesium the use of these materials as primary ingredients in the process is reduced to a very small amount and, in effect, electric power is substituted for them.

The advantages of the process outlined above are made evident by a comparison with the processes of the prior art. .The production of beryllium has heretofore been attempted by reducing a beryllium halide with some of the metals of the alkali and alkaline earth groups, but because of the volatility of the beryllium halides and their great tendency to oxidize, such a reduction must be carried on in a tightly closed strongly built vessel or bomb, and for commercial size operaltions such a scheme is impractical. Also the metal produced is predominantly fog, and much of its oxide. The use of other fused salts as a diluent for the beryllium salt makes the closed vessel unnecessary and permits of a controllable 3 reaction on a large scale at comparatively low temperature. It also protects the beryllium salts from oxidation and produces very pure metal. The magnesium replaces the beryllium slowly and quietly, but very completely, permitting practil cally full recovery of the contained beryllium, and in a useful form rather than as fog. A further improvement is brought about by the use of beryllium chloride, which produces a water soluble end product. This renders the separation of the beryllium easy and emcient.

Magnesium has heretofore been used as a reducing agent for beryllium only from its fluoride, either to produce magnesium-beryllium alloy, or pure beryllium. This produces the insoluble magnesium fluoride, hence the beryllium metal cannot be separated by leaching, and the reaction must therefore be carried on at a temperature above the melting point of beryllium. The difliculties of finding suitable vessels for holding fluoride fusions at such temperatures are very great, and the loss of beryllium fluoride by vaporization is also very serious. The metal produced must then be re-melted one or more times for purification, with serious losses at each melting. These operative difliculties and low recoveries are overcome in the preferred process of this invention.

While the preferred method of practicing the present invention has been specifically described, it is to be understood that the same is merely illustrative of the general method, and that the invention is not limited thereto, but is obviously capable of many variations, and I particularly point out and claim as my invention the following:

1. The process of making substantially pure beryllium metal which comprises making a fused mixture of sodium chloride, potassium chloride and beryllium chloride, and introducing metallic magnesium into this mixture.

2. The process of making beryllium metal which comprises making a fused mixture of sodium chloride; potassium chloride and beryllium chloride in which the beryllium chloride is approximately from flve percent to seventy-five percent of the total mixture with the remainder substantially equal parts of sodium chloride and potassium chloride, introducing metallic magnesium into the mixture, leaching the resultant product with water, and separating out the beryllium metal.

3. The process of making beryllium metal which comprises making a fused mixture of sodium chloride, potassium chloride and beryllium chloride, in which the beryllium chloride is approximately twenty-flve percent of the total mixture with the remainder substantially equal parts of sodium chloride and potassium chloride, introducing metallic magnesium into the mixture, leaching the resultant product with water, and separating out the beryllium metal.

4. The process of making beryllium metal which comprises making a fused mixture of salts one of which is beryllium chloride and the remainder of which is water soluble and does not react appreciably with magnesium, introducing metallic magnesium into the mixture, leaching the resultant product with water, and separating out the beryllium metal.

5. The process of making beryllium metal which comprises making a fused salt mixture containing beryllium chloride, introducing metallic magnesium in a form having small crosssection into the mixture while the same is continuously stirred, leaching the resultant product with water, and separating out the beryllium metal.

HUGH S. COOPER.