US2817630A - Methods of producing titanium and zirconium - Google Patents

Description

METHODS OF PRODUCING TITANIUNI AND ZmCONIUM Reginald 5. Dean, University Park, Md., assignor to Chicago Development Corporation, Riverdale, Md.

No Drawing. Application February 4, 1954 Serial No. 408,310

1 Claim. (Cl. 204-64) This invention relates to the production of titanium and zirconium. It has for its object the electrolytic production of pure titanium and zirconium.

In the known art, titanium and zirconium are electrorefined in several fused salt electrolytes. In electrolytes of alkali metal or/ and alkaline earth metal, or magnesium, chlorides, bromides and iodides, such electro-refining can be carried on with oxygen-containing electrodes, the oxygen therein remaining in an anode residue. in such electrolytes the mechanism is that stated, e. g. by Van Arkel in the text Reine Metalle, Berlin, Jul. Springer (1939): The melt always contains as positive ions practically only alkali and alkaline earth ions. These carry the electricity and are discharged at the cathode. The alkali and alkaline earth metals then reduce the higher valent halogenide. The metal that is thus formed separates not as compact masses on the cathode but forms smaller or larger crystals throughout the melt. In such electrorefining it is necessary to separate anode and cathode areas, and to prevent so far as possible codeposition of the alkali metals, alkaline earth metals or magnesium with the titanium or zirconium. This separation of the metal in the electrolyte rather than on the cathode has presented a problem recognized by many investigators. For example, Steinberg et al. writing on Extractive Metallurgy of Zirconium, in Zirconium and Zirconium Alloys, Amer. Soc. of Metals (1953) say: By utilizing current densities of 5 amperes per square centimeter or less it is possible to obtain plate like deposits on the cathode; however, these have in the past also been contaminated with salt. At higher current densities dentrites are obtained and at much higher current densities black amorphous zirconium powder permeates the bath. This separation has been accomplished in various ways.

1 have found that the desired results can be obtained by observing continuous depolarization of the cathode. The term depolarization is here used in its general sense as defined, for example, in the text of Elektrochemie Geschmalzener Salze, by Paul Drossbach, Berlin, Springer (1938) at page 55: The polarization voltage is only equal to the EMF of the cell when the element is reversible. In general it is greater or less than the EMF, and we call the difference between the polarization voltage and the EMF either over-voltage or undervoltage. For the term over-voltage one finds frequently the term Polarization, which is not to be confused with polarization potential. Conveniently, I place in the cell a reference (or auxiliary) electrode of the pure metal which it is desired to deposit, and I add at the cathode a suitable depolarizing agent, e. g., chlorine or titanium tetrachloride or zirconium tetrachloride at a rate to maintain the voltage between cathode and reference electrode at practically zero, e. g., at not substantially more than 0.05 volt. The amount of such depolarizing agent used is a significant amount not in excess of by weight of the stoichiometric equivalent of the Ti or 2,817,639 Patented Dec. 24, 1957 Zr being deposited. The measurement of polarization in a fused salt cell having anode and cathode of the same metal in an electrolyte containing alkalinous metal chloride is described by Aten, Hertog and Westenberg in a paper published in Transactions of the American Electrochemical Soc., 47 (1925), page 265, wherein the authors say, in part: A tube of pyrex glass, drawn to a point, was placed in the fused salt close to the cathode and a wire of the metal being deposited was placed in this tube. The voltage between polarized cathode and this auxiliary electrode was measured with a simple potentiometer.

By the term depolarizing agent I mean to define any substance which when added at the cathode will bring the potential of the cathode to that, or to substantially that, of the reference electrode. This result is depolarization as the term is used in the references cited. I believe that the depolarization of the cathode is caused by reaction with the alkali or alkaline earth metal deposited there. This conclusion is consistent with the generally accepted fact of their deposition in cells having alkalinous metal chloride electrolytes; and, since the alkalinous metals are'known (Quill, Thermodynamics of Miscellaneous Materials) to dissolve in their fused halides, their diffusion away from the cathode to combine with the anode product is highly probable as stated by Drossbach in his text previously cited, page 65, the metal solution formed at the cathode moves towards the anode and reacts with the anode product. This reaction taking place in the electrolyte obviously tends to prevent the anode product from reaching the cathode by diffusion. Accordingly, the function of the chlorine, or of the tetrachloride of titanium or zirconium introduced immediately adjacent the cathode, is to promote transport of titanium, or zirconium, in the bath itself. Observance of this measure substantially prevents co-deposition, on the cathode, of magnesium and/or alkali metal and/or alkaline earth metal along with the titanium or zirconium.

I prefer to use a porous cathode of graphite or titanium or zirconium through which porous cathode the chlorine gas-or the titanium tetrachloride or zirconium tetrachloride vapor--is introduced into the electrolytic bath at controllable rate.

Under these conditions the metal-titanium or zirconium, as the case may be-is deposited in compact form (not a plate) on the cathode, which latter may be removed from the cell with its deposit attached. Apparatus for protecting the bath and removed electrodes is conventional. The cathode, after removal from the cell, is stripped of its deposit, and the latter may be treated in known ways for the recovery of pure metal.

The anode for use in electrorefining by the present process may be a titanium-oxygen or a zirconium-oxygen alloy containing up to 10% oxygen, produced in known ways. I prefer, however, to use an anode prepared by reacting pure titanium oxide, or zirconium oxide, with magnesium. This latter reducing agent is selected because it does not readily form titanates and does not form a carbide.

I prefer to carry out this reduction in such a way that the reduced titanium, or zirconium, may be separated in considerable measure from residual magnesium and magnesium oxide. This is conveniently accomplished by triturating the titanium or zirconium oxide under a bath of molten magnesium. The magnesium oxide formed being lighter than the metal floats to the surface of the molten magnesium and the reduced titanium, or zirconium, is wetted by the magnesium. The titanium or zirconium sinks to the bottom of the bath and may be recovered therefrom in various ways known in the art. I may, for example, separate the titanium and liquid magnesium by filtration or centrifuging or by mechanical treatment after solidification. The product will contain up to magnesium.

Such anode material may be compacted, by known procedures, into anodes which are electrically conducting. A preferable procedure is to mix the comminuted anode material with a small but effective amount of a tar or pitch binder, and use the mixture as a self-baking, or Soderberg, electrode. In use, the magnesium oxide and any carbon which may be contained in this anode fall to the bottom of the cell from which they are pen'odically recovered.

Any magnesium in the anode enters the electrolyte as magnesium chloride. For this and other reasons, I prefer to use a magnesium chloride electrolyte. In this way the electrolyte content of the cell may be maintained directly and simply. While magnesium chloride is the preferred electrolyte, other halide salts of alkali metals and/or alkaline earth metals may be substituted therefor.

In order to more fully distinguish my invention from the prior art, the following discussion is given.

I am familiar with the old art of reducing titanium and z'ponium oxides with magnesium. Such processes are disclosed, for example, in U. S. Patent 1,602,542, and more recently in British Patent 675,933, of July 16, 1952. In this prior art, the magnesium is pulverized and intimately mixed with the oxides. No more than a slight excess of magnesium is used. The product is an intimate mixture of the metal to be reduced, magnesium oxide, and excess magnesium. This product has been treated with acid to remove magnesium and magnesium oxide. This procedure is expensive and furthermore yields an impure metal. The addition of calcium chlo ride to the reacting mixture has been recommended, but brings about only slight improvement. My process is based on the conception that the reaction between titanium or zirconium dioxide and magnesium is not reversible, but proceeds in one direction to produce magnesium oxide and pure titanium or zirconium without intermediate oxides. The reverse, however, forms intermediate oxides such as TiO. It is therefore necessary to remove the magnesium oxide from the reaction zone as formed. The addition of fluxes, as in the known art, is inadequate. I have found, however, that if the reaction is. carried on under a bath of molten magnesium, the products of reaction will be immediately separated by virtue, of their diflYerent surface and gravity relationships to the bath. The magnesium oxide is not wetted and will float to the surface. The reduced metal is wetted and sinks. My invention, therefore, is based on the novel conception of the reaction mechanism, and is characterized. by the procedure just described.

It will be clear that my procedure for reducing titanium and zirconium oxides produces an ideal product for elec trorefining since the principal impurity is metallic magnesium. This may be removed by sublimation, but it is preferably left in part in the anode whereby to supply make-up magnesium chloride in the electrolyte. When this is done enough chlorine or titanium tetrachloride or zirconium tetrachloride is added to combine with the excess magnesium. This material is added at the cathode, and only in such amounts as to prevent deposition of magnesium on the cathode. I am familiar with processes in the art in which titanium tetrachloride is added at the cathode for reduction to titanium. My process differs from these in the controlled addition. In the known art no such control is provided and the result is the deposition of magnesium along with the titanium.

Example I I take an anode of titanium containing approximately 8% of oxygen, in the form of a titanium-oxygen alloy, and place it together with a porous graphite cathode in an electrolytic cell having a bath of molten magnesium chloride electrolyte maintained at about 850 C. I- pass a unidirectional current between the electrodes at about 1000 amp./sq.ft. of efiective cathode surface, and pass titanium tetrachloride through the porous cathode into the electrolytic bath at such an amount-about 5% of the stoichiometric equivalent of the titanium being deposited-that the cathode is maintained at substantially zero potential against a reference electrode, of pure titanium, likewise positioned in the bath of the cell.

Pure, or substantially pure, titanium is deposited on the cathode as a coarsely crystalline, compact mass which adheres to the. cathode sufficiently that it may be lifted from the cell bath with the cathode.

Example 11 In this example I take titanium dioxide and introduce it into a bath of molten magnesium at 900 C. I separate the titanium rich layer which forms at the bottom of the: bath. I compress the resulting product, which contains approximately titanium, together with magnesium and a small amount of magnesium oxide, into an anode and use it as in Example I except that I use chlorine for depolarizing the cathode. To the cathode there attaches a coarsely crystalline, compact deposit of substantially pure titanium. After suitable prolongation of the electro-refining operation the cathode with deposit attached may be raised above the electrolyte bath and the deposit thereafter stripped from the cathode.

Example 111 In this example I take hafnium-free ZrO and mix it, as in Example II for TiO into a bath of molten magnesium at 900 C., separate the zirconium rich layer, compact the resulting product, which is a mixture of about 90% zirconium with magnesium and a little metallic magnesium oxide, into an anode which latter I use as in Example I. To the cathode there attaches a coarsely crystalline, compact mass of substantially pure zirconium, which may be stripped from the cathode after the latter has been lifted from the bath.

Example IV I proceed as in Example II except that instead of compressing the magnesium-reduced product I mix it with approximately 5% pitch and use it as a self-baking anode in an electrorefining procedure like that of Ex ample I using titanium tetrachloride for depolarizing the cathode. The anode residue is found to contain the carbonaceous residue of the pitch content of the anode together with the magnesium oxide.

The titanium deposit on the cathode is substantially identical with that produced in Example I.

Example V I proceed as in Example II except that I use as electrolytev the eutectic mixture of lithium and potassium chlorides, and maintain the electrolytic bath at a temperature of about 500 C. The metal deposit on the cathode is more finely crystalline and more spongy than in Example I, but nevertheless is adherent to the cathode and removable with the latter from the bath. The strippable product is substantially pure titanium.

Example VI electrolyte other, per se known molten electrolyte baths composed of mixtures of bromides, or iodides, or preferably chlorides of the alkali metals and/ or alkaline earth metals or magnesium.

Example VII I take pure TiO and introduce it into the bottom of a molten bath of magnesium at 900 C. I mix the oxide with the magnesium by trituration. I recover a product from magnesium which analyzes approximately 90% titanium, containing 5% oxygen, magnesium. I compress this material into an anode and use it as in Example I with similar results.

Example VIII I proceed as in Example VII, and recover a product from the lower part of the magnesium bath which contains about 80% titanium and magnesium. I commiute this material in a hammer mill and screen it on a 20 mesh sieve. The magnesium is flattened and will not pass through the sieve. The titanium being brittle, passes through the sieve in a product containing 95% titanium and 5% magnesium.

Example IX In this example I proceed as in Example VIII to obtain the mixture of titanium and magnesium. I compress this mixture into a form suitable for anodes. I then heat these anodes in a vacuum to 1100 C., and sublime the magnesium from them. The resulting product, containing about 1 percent magnesium, is used directly as an anode for electrorefining.

I claim:

A process for electrorefining a metal of the group consisting of titanium and zirconium, which comprises the steps of passing a unidirectional direct current from an anode, consisting essentially of said metal containing oxygen in interstitial solid solution to an inert solid cathode through a molten bath of at least one member of the group consisting of alkali and alkaline earth metal chlorides initially free from dissolved salts of said metal, simultaneously introducing into the bath adjacent the cathode a depolarizer selected from the group consisting of chlorine and the tetrachloride of said metal, said depolarizer being added at such a rate and in such an amount, not exceeding about 10% by weight of the stoichiometric equivalent of the metal being produced, as to maintain a cathode potential against an auxiliary electrode of said metal in pure form of not more than 0.05 volt, whereby to produce substantially pure metal in coherent form on said cathode, removing said cathode and adherent metal from the bath, and cooling said cathode and adherent metal in a neutral atmosphere.

References Cited in the file of this patent UNITED STATES PATENTS 2,537,068 Lilliendahl et al J an. 9, 1951 2,707,169 Steinberg et al Apr. 26, 1955 2,714,575 Wainer et al. Aug. 2, 1955 2,722,509 Wainer Nov. 1, 1955 2,731,404 Wainer Jan. 17, 1956 2,734,856 Schultz et a l Feb. 14, 1956 2,749,295 Svanstrom et al. June 5, 1956 2,755,240 Normore et a1 July 17, 1956 FOREIGN PATENTS 682,919 Great Britain Nov. 19, 1952 OTHER REFERENCES Journal of Metals, September 1956, pp. 1165-1166.

Mantell: Industrial Electrochemistry, first edition 1931 pp. 46 and 47.

U. S. Bureau of Mines, Report of Investigations RI 4915, November 1952, pp. 17 thru 26.