Production of subdivided metals

United States Patent Office ²’⁸³¹’⁸⁰²

Patented Apr. 22, 1958

2,831,802

PRODUCTION OF SUBDIVIDED METALS

Ben B. Raney, Linton, Ind., assignor to Chicago Development Corporation, Riverdale, Md.

No Drawing. Application November 14,1951 Serial No. 256,385

Claims. (Cl. 204—10)

This invention relates to the production of heavy metals in subdivided form.. It relates particularly to processes for tire reduction of heavy metal compounds to metallic particles by reacting these compounds with solutions of alkali or alkaline earth metals in fused chlorides of those metals. I have found that such reduction is more rapid than reduction by alkali or alkaline earth metals in other forms such as massive or vapor states. It also produces the heavy metals in the form of characteristic particles depending on the compound reduced. If the compound is solid and insoluble in the molten chlorides, the*metal particles are pseudomorphic after the compound. If the compound is soluble in the melt, the metal particles will be filamentary.

By heavy metals, I mean to refer in this specification to the metals in subgroup A of group IV, i. e., the metals of the. titanium group of the periodic system, titanium, zirconium, and hafnium. For convenience, I will refer to the alkali and alkaline earth metals as alkalinous metals.

To carry out my invention the chlorides are fused and saturated with alkali or alkaline earth metal, the solution of the metal in the absence of undissolved metal is reacted with the heavy metal compound to be reduced. The heavy metal compound must not, however, be introduced at. a greater rate than the fused salt solution is replenished with alkalinous metal.

This is conveniently done by introducing the alkalinous metal into the stirred bath only as rapidly as it is dissolved and introducing the heavy metal compound in disperse form at the same rate in another portion of the bath.

I have found that in order to dissolve the alkalinous metal in the fused chlorides they must be substantially free from dissolved oxide. The process of my invention is inoperative if there is a substantial amount of oxide dissolved in the bath. The amount of dissolved_: oxide in fused commercial salts is too great. Several means of deoxidizing the bath are set forth in a later section. In certain instances, the subdivided metal to be produced may deoxidize the bath. This, however, is undesirable as the subdivided metal will be contaminated with oxide which is not readily separated. My invention, therefore, contemplates a bath which is deoxidized in some manner so that the products of deoxidization may be readily separated from the subdivided metal.

In my co-pending application, Serial Number 152,175, filed March 27, 1950, and my co-pending application, Serial Number 230,336, filed June 7, 1951, now abandoned, I have disclosed the preparation of titanium powder. The methods disclosed in these applications differ from the present method in many important respects although a superficial similarity exists. In the methods of these applications no provision is made for deoxidization of the bath or for the conditions necessary to obtain a solution of the alkalinous metal in the fused salt.

The heavy metal compound may be a chloride or oxide. The compound to be reduced and the alkalinous metal must be so selected that their reaction at the temperature Of the fused bath has a negative free energy. Data for ascertaining this point will be found in the literature (see U. S. Bureau of Standards Circular No. 500, February 1, 1952) jmd for purposes of applying it to the present invention, the activity of the alkalinous metal in solution may 5 be considered unity.

A preferred form of carrying, out my invention consists in the reduction of heavy metal chlorides or oxides to .rnely divided metal by bringing them into the solution of alkalinous metal in fused chloride produced by electrolyte sis of oxide free alkali and alkaline earth metal chlorides, at a place away from the cathode and at a suitable rate so that the alkalinous metal solution is always in excess, but no undissolved alkalinous metal is present at the place of reaction.

The anode in this embodiment of my process may be graphite in which case chlorine is evolved and removed from the cell by suitable means. It is frequently convenient however, to provide the heavy metal compounds to be reduced by anodic reaction. In the simplest in20 stance, these compounds may be merely the chlorides of the anode metal or metals. The anode may advantageously be made, however, of partially reduced metal in which case, oxides and chlorides are provided by the anode. It is necessary that there be enough metal in such 25 partially reduced material that the anode be conducting.

Other conducting heavy metal compounds which form chlorides when used as anodes in the molten chloride bath may be used as anodes.

• Anodes which react with chlorine to form chlorides 30 such as mixtures of the heavy metal oxides and carbon may also be used.

I am familiar with the reduction of metal compounds by adding them to alkalinous metals in the presence of fused salts. Such procedure is not in any way equivalent 35 to my invention. When the compound to be reduced comes, in contact with the alkalinous metal in the mass, there is no solution of the alkalinous metal in the fused chloride. The compound is imperfectly reduced and the resulting metal is usually globular or sponge-like and not 40 in discrete particles in the form described for my invention.

I am also familiar with the known art of producing subdivided metals by the electrolysis of molten electrolytes containing salts of these metals in solution. Such 45 Procedure is well known, for example, for the production of powdered iron. In such processes, the metal is deposited in idiomorphic crystals which may form nodular aggregates, but not as pseudomorphs or filaments as in my process.

. I am also familiar with the deposition of metals like titanium, zirconium, and tantalum from the fluotitanates, fluozirconates and fluotantalates respectively in which the metal is also formed at the cathode.

I am also familiar with attempts to reduce compounds _5g insoluble in molten electrolytic baths by introducing them at the cathode and bringing about a reaction with the alkali or alkaline earth metal formed. An example of this procedure is found in British Patent 1,759 of 1904 issued to Borchers et al. In this instance, titanium di60 oxide is introduced in excess at a eathode in a bath containing calcium, chloride and the titanium dioxide is in some measure reduced by the calcium formed.

This procedure is unsatisfactory because the compound to be reduced being present in excess is not completely re₆₅ duced. Furthermore, and this is an important distinction of the prior art from my process, reduction in the processes of the art is brought about by discrete masses of the metal which is less effective than reduction by a solution of the jnetal in the fused chloride. It will be seen 70 that if there is any reducible heavy metal compound at the cathode, there will be no opportunity for the formation of the solution of the alkalinous metal in the salt bath which

2,831,802 solution is essential to my invention. Accordingly, it is of the essence of my invention that the heavy metal compound to be reduced be introduced in disperse form at a place away from the cathode and at a rate no greater than it is reduced to metal.

It is also Of the essence of my invention that the heavy metal compound, chloride or oxide, be selected with respect to the fused chloride bath so that the free energy of the reaction for the reduction of the compound to metal by the alkalinous metal of the bath be negative at the temperature of electrolysis. These free energies may be found in many publications.

From these data it will be seen for example that all the alkali and alkaline earth metals reduce heavy metal chlorides, but that TiO₂ and ZrO₂ are not reduced by Na or K, but are reduced by Ca. Hence if it is desired to reduce TiO₂ or ZrO₂ to metal by my process, the electrolyte should contain CaCl₂.

It will be noted further that when reduction of chlorides is carried out, the bath is simply replenished in its alkali or alkaline earth chloride content. However, when heavy metal oxides are reduced, an oxide of an alkali or alkaline earth metal is formed. When this is an alkali metal oxide, it must not be allowed to accumulate in the bath in substantial concentration as it will react with and oxidize the subdivided metal. The bath must be frequently changed and deoxidized by methods which will be described in detail hereinafter. If the reduction is by an alkaline earth metal then the reduced metal will be contaminated with insoluble oxide of that metal and separation must be subsequently made.

The shape and form of the metal formed by my process is important because in the production of pure metals, I may make use of this shape factor to separate the reduced metal from anode fragments of the Original metal and from non-metallic anode slime or other non-metallic materials. The methods of separation of solids based on shape factors are well known and include gravity and electrostatic methods. For example, the bundles which form from the filamentary particles are easily washed over the riffles of a shaking table, whereas the more symmetrical particles fall into the riffles of the table.

In other cases, the pseudomorphic particles of metals may be of other shapes but always different from the anode fragments and slimes. Separation of the newly formed metal particles from the anode fragments and slime may be made by other processes such as flotation in which size, shape, and surface condition are all factors.

The conditions of electrolysis are not particularly critical. Current densities at the anode may be 100-5,000 amperes per square foot, cathode current density may be 100-1,000 amperes per square foot. The anode may be of the impure metal it is desired to produce in pure subdivided form. The cathode may be any unattacked metal or carbon. When the anode is of the impure metal it is desired to produce, the cathode may be the same metal and the current may be from time to time reversed in order to produce subdivided metal from both electrodes. In this case the vessel holding the bath may be any unattacked metal or ceramic. When using a carbon anode, the cathode may conveniently be the metal vessel which contains the salt.

My method for the production of pure metals in solid subdivided form is applicable to those metals which are not molten at the necessary temperatures of operation, minimum about 450° C., and which do not alloy with the alkali or alkaline earth metals used and which meet the requirements previously set forth concerning die free energy of their compounds.

The process is applicable to mixtures of compounds of several metals. Such mixtures of compounds can be supplied in various ways, but conveniently from an alloy anode. Such mixtures of compounds are reduced to discrete unalloyed particles of the metals. Such mixtures of metal particles can be separated by mechanical or chemical means.

As previously mentioned, the oxygen content of the bath must be kept low to obtain metals relatively free of 5 oxygen. This may be accomplished by using a very large amount of oxygen-free molten chlorides with relation to the metal deposited and removing the oxygen from the chlorides before recirculating them in the system. The oxygen can be removed from molten chlorides by pass10 ing hydrochloric acid gas through them or by treating them with calcium metal. This method involves the separation of the metals from the salts before the latter can be deoxidized. This may be done by solution of the. salts in water, or by separating the molten salts mechanically 15 from the metals. Gravity or hot pressing may be employed. I prefer, however, to deoxidize the molten salts without separating them from the subdivided metal. I accomplish this in one of several ways. In one example I react the salt with incandescent carbon. The re20 action of K₂O, for example, with incandescent carbon is as follows:

K₂O+C-^C0+2K

The incandescent carbon is maintained in a limited 25 zone of the salt bath for example an electrically heated graphite spiral. The potassium vaporized at the graphite is condensed in the bath away from the spiral, while the CO escapes. In some cases a small amount of carbide metal may be formed by reaction of the incandescent 39 carbon with the subdivided metal. This adheres to the spiral. The amount of carbide formed is small compared to the carbon monoxide because the oxygen being in solution has much higher mobility and diffuses from throughout the bath to react while only a local zone of subdivided 35 metal reacts.

In another method of deoxidizing the salt bath, I use 1 heated filament of metallic iron. This reacts with the alkali metal oxide in accordance with the reaction:

4M₂O+3Fe-> Fe₃O₄+8M

The alkali metal is vaporized away from the iron and is dissolved in the fused salts. With these procedures, there may be some reaction of the subdivided metal and the Fe₃O₄ to produce oxidic compound which, however, j.-; adhere to the iron filament and do not contaminate the remaining subdivided metal.

Other reactions may be employed in which the oxygen of the bath is removed by a reaction which depends on its mobility and hence the product is concentrated locally, ,; and there is a minimum of reaction with the subdivided metal.

I prefer to use in my process the lowest melting mixture of alkali and alkaline earth metal chlorides which is consistent with the free energy requirements for reduc55 tion which have been set forth. In order to reduce oxides, I prefer to have at least one alkaline earth chloride present since the negative free energy of reduction of heavy metal oxides by the alkaline earth metals is greater than for the alkali metals. In fact, several 50 metals which are utilized for the practice of my invention as titanium and zirconium are not reduced from their oxides by sodium and potassium.

I have described my invention with the use of chlorides of the alkali or alkaline earth metals because they are c; practically advantageous. Bromides or mixtures of bromides and chlorides may be used without departing from the spirit of my invention and are considered the chemical equivalent of the chlorides. Fluorides may also be added in small amounts to remove and volatilize sili70 con, or to render alumina present in the heavy metal compounds soluble in the bath. Borates may be added to dissolve alkaline earth metal oxides to prevent their contaminating the metals produced. Such additions to the chloride bath .are encompassed within my invention.