US2752303A

US2752303A - Fused bath electrolysis of metal chlorides

Info

Publication number: US2752303A
Application number: US453898A
Authority: US
Inventors: Hugh S Cooper
Original assignee: Individual
Current assignee: Individual
Priority date: 1954-09-02
Filing date: 1954-09-02
Publication date: 1956-06-26
Anticipated expiration: 1973-06-26

Description

June 26, 1956 H, 5, COOPER 2,752,303

FUSED BATH ELECTROLYSIS OF METAL CHLORIDES Filed Sept. 2, 1954 INVENTOR Hugh S. Cooper BY c www +72( L? ATTORNE Unite FUSED BATH ELECTRLYSIS F METAL CHLORHDES Hugh S. Cooper, Shaker Heights, Ohio, assigner' to Walter M. Weil, Cleveland, Ghio This invention relates to electro-metallurgy and particularly to electrolytic methods or processes for produc` ing substantially pure metals by the fused bath electrolysis of their chlorides, and to apparatus for carrying out such methods and processes.

The principal object of the invention is to provide a process suitable for general application to the fused bath electrolysis of anhydrous chlorides of a number of metals, particularly those that are difficult to produce economically in pure elemental form by other processes.

A more specific object of the invention is to increase the efficiency of processes involving the fused bath electrolysis of metal chlorides generally, thereby rendering such processes commercially practical for the first time, when applied to `a number of metals, and improving the economy of those processes of this character which have been commercially practiced heretofore.

Another object of the invention is to provide a novel form of apparatus particularly suited for use in electrolyzing fused metal salts in accordance with the invention.

A number of metals, notably, magnesium, beryllium, cerium, thorium, and the like have been commercially produced by the fused bath electrolysis of their chlorides. Various bath solvents have been employed, including alkali metal chlorides, which acted to prevent decomposition of the chlorides to be electrolyzed and to maintain the baths in a properly iluid condition at reasonable operating temperatures. ln the case of beryllium, for example, the chloride is extremely hygroscopic and unstable in air at elevated temperatures, and protection of beryllium chloride from oxidation while being added to and melted in the fused salt bath has been accomplished by maintaining a reducing atmosphere of hydrogen over the bath, as disclosed in my prior U. S. Patent No. 1,805,567.

Upon `attempting to apply the same general process to the electrolysis of the higher chlorides of certain polyvalent metals, particularly chromic chloride, ferric chloride, and mixtures thereof, reasonably satisfactory operation on `a small scale was achieved when proper bath compositions and equipment were employed. However, when an attempt was made to carry out the same process on a substantially larger scale, as required for commercial operations, great diiculty was encountered. The metal recovery dropped from around 90% down to around 50%, the current efficiency was substantially reduced, even when calculated on the basis of the reduced yields, and it was impossible to exhaust the bath of the material 4being electrolyzed. This last effect is particularly serious in the case of chromium, for example, since chromium chlorides remaining in the bath are entrained with the product and decompose rapidly in the leaching process to form CrzOs. The Cr2O3 cannot be separated from the metal by any practical purication process and greatly reduces the value and usefulness of the product.

In an effort to solve these perplexing problems in the electrolysis of chromium chloride, I employed an eX- atent pedient that contributed materially to the success of the process as applied to chromium and to various other metals las well. This discovery has not only proved to be of great value when applied to the fused bath electrolysis of chlorides of polyvalent metals, such as chromium, iron, manganese, nickel, cobalt, tantalum, columbium, uranium, molybdenum, tungsten, etc., but it also contributes to the eiiicient fused bath electrolysis of chlorides of monovalent metals, such as aluminum, beryllium, magnesium, thorium, zirconium, etc. Thus, in addition to facilitating the economical production of certain polyvalent metals by fused bath electrolysis of their chlorides, this expedient also improves the current eiciency and yields when producing monovalent metals by the same general process. Since my process involves the employ ment of a substantially pure alkali metal chloride bath as a solvent for the metal chloride or metal chlorides to be electrolyzed, it is limited in its application to metals below such alkali metals in the electromotive or electrochemical series, and preferably below calcium.

Broadly stated, the expedient referred to above involves the forcible diffusion of hydrogen gas throughout the fused bath, and particularly around the anode surfaces where chlorine is released during electrolysis. This diffusion of hydrogen gas is preferably and most conveniently accomplished by feeding the gas downwardly through a hollow anode suspended in the bath and discharging the gas into the bath through apertures at or adjacent the lower end of the anode and adjacent the bottom of the bath. The discharge apertures are preferably directed generally, or in part, in a downward direction so that the gas reaches the very bottom of the bath and performs a stirring or agitating function as it bubbles upwardly tosthe surface of the bath.

The effect of the forcible diffusion of hydrogen gas through the bath is most pronounced when electrolyzing chlorides of polyvalent metals, and its operation will be explained by further reference to the electrolysis of chromium chloride. When chromium chloride is employed as the source of chromium, the electrolytic reduction of the chromium appears to involve successive steps in which the chromic chloride is first reduced to chromous chloride, followed by reduction of the chromous chloride to elemental chromium. Both steps involve the release of chlorine at the surface of the anode. By reason of the great activity of the free chlorine thus released within the bath, the electrolytic reduction steps appear to be reversed by chemical recombination of chlorine with elemental chromium deposited at the cathode and with chromous chloride in the bath, the elemental chromium being converted back to chromous chloride and chromous chloride in turn being reconverted to chromic chloride. This continues at such a rate that it is impossible to deplete the bath of chromium chloride, and thus many times the theoretically required current is consumed for each gram of chromium metal eventually recovered from the cell.

If the process is performed by employing chromous chloride initially as the source of chromium, instead of chromic chloride, the same problem is encountered to substantially the same degree, with some chromous chloride actually being converted to chromic chloride along with the conversion of deposited chromium metal back to chromous chloride.

Because of the intense ainity of hydrogen for chlorine, the hydrogen forced throughout the bath in accordance with the present invention not only acts to take up the electrolytically released chlorine, forming HCl, and to perform a valuable agitating function, but it also acts chemically and/ or thermically to reduce chromic chloride to chromous chloride and chromous chloride to metallic chromium. This reduction supplements the electrolytic reduction and may actually result in some instances in an overall current eiiciency greater than theoretical, as will be shown by specific examples given herein.

From my observations of the effectiveness of hydrogen gas in the bath as a chemical and/or thermic reducing agent, as well as an agent for tying up. the released chlorine, when applied to the electrolytic reduction of chro miurn chlorides, I concluded that its use in the reduction of other metal chlorides would be similarly beneficial, even where reasonably satisfactory commercial scale op erations have been feasible without employing such an expedient. As` will be shown hereinafter, this has proved to be true, and numerous different metal chlorides may be reduced electrolytically in accordance with the invention with extraordinarily high yields of pure metals and with an actual, or apparent, current eiiiciency of close to 100%, or better than 100% in some instances.

The expedient of forcibly diffusing hydrogen gas through the fused bath contributes substantially to success of the process, as explained above, and has been found to be essential to the economical production of many polyvalent metals by fused bath electrolysis of their chlorides. However, in the course of employing that expedient in the process, first on a small laboratory scale and then on progressively larger and larger scales, l discovered that other diiiiculties arise as the size of the electrolytic cell is increased. When employing large electrolytic cells containing of the order of 100 lbs. or more of the fused bath, the process, which worked so well on a smaller scale after adopting the above described expedient, failed to produce the pure product so necessary for most commercial uses and suffered a serious drop in current efficiency.

One phenomenon responsible for this failure of large scale operations is the formation in the bath of large amounts of oxide of the polyvalent metal being electrolyzed. This oxide remains and accumulates in the bath as electrolysis is continued. Entrainment of this oxide with the deposited metal renders it virtually impossible to recover a product free from substantial amounts of the oxide.

Another phenomenon which contributes to this failure is an apparent reconversion of deposited metal back to its chloride, rendering it impossible to exhaust the bath and reducing the overall current efficiency of the process.

The exact explanation of these phenomena still remains obscure. However, I have found that the additional difficulties caused thereby may be eliminated by employing a special form of pot to hold the fused bath and to serve as the cathode of the cell, and by manipulating the process in the pot in a special manner. The special form of pot is one made of a suitably corrosion resistant alloy, such as Inconel (a nickel-chromium-iron alloy containing about 'l0-75% nickel, about 15% chromium, and the balance principally iron) and having a non-metallic, electrically non-conductive, refractory, upper inner surface about the entire periphery of the pot. Suitable refractories for this purpose are magnesia, zirconium silicate, zirconium oxide, calcined dolomite (magnesium and calcium oxides), aluminum silicate, and the like, magnesia being preferred. Such refractory materials are highly resistant to chemical attack by chlorine, hydrogen chloride, and the other contents of the pot. The special manipulative procedure involves maintaining the fused bath level above the lower edge of this refractory surface at all times during electrolysis of the chloride of the metal to be produced, whereby the plane of the bath surface always intersects the refractory wall of the pot around the entire periphery.

The special pot and manipulative operation described above somehow act to prevent the formation in the bath of metal oxides which are water and acid insoluble and virtually impossible to remove from the metal product by chemical treatment. Because of the non-conductive character of the refractory surface, no metal is deposited near the surface of the bath, where the opportunity for its oxidation and rechloridization by gases in the cell may be most pronounced. This may be one important factor in the solution of the former difficulties. By the same token no metal of the pot is exposed adjacent or above the surface of the bath, where the opportunity for its attack may be greatest. This may be another contributing factor in the solution of the former difculties. Whatever the full explanation may be, these features of the present invention are effective to substantially completely prevent oxide contamination of the bath and of the product, enable the bath to be substantially completely exhausted of the metal to be produced, and restore the high current efficiencies previously obtainable only with cells too small for significant commercial production.

A more complete understanding of the invention and the problems which it has overcome will be obtained by reference to preliminary experiments in which chromic chloride was electrolyzed in a fused bath of alkali metal chloride on a small scale, and by specific examples of the behavior of the process when applied to large scale oper?- tions, both with and without employing the present invention. rThe preliminary experiments were carried out in apparatus consisting of a simple graphite lined crucible capable of holding a bath of about 9 to l0 kgs. of fused metal salts and forming the anode of the cell. The cathode consisted of a rectangular plate of Inconel suspended from a rod of nickel. The chromium chloride used as a source of chromium was pressed into sticks or rods, and these were broken into pieces about walnut size at the time of use in order to minimize conversion of the chloride to oxide by atmospheric oxygen and to minimize the absorption of moisture from the air.

Quantities of chromium chloride of about 500 gms. were melted in a bath of about 5 to 6 kgs. of either potassium chloride, sodium chloride, or a mixture thereof, at temperatures ranging from 750 to 900 C. Upon applying in a potential of about 7 volts across the anode and cathode of the cell, currents of about 500 amperes were obtained, and the bath remained clear and quiet during electrolysis. Continuance of electrolysis until from to 200% of the current theoretically required to deposit all of the chromium has been passed through the cell produced between about 5 to 45% yields of chromium of only fair purity after washing with water and acid. Chlorine gas liberated from the anode severely attacked the heated portions of the cathode and its support, destroying these parts at an excessive rate and causing contamination of the metal deposited on the cathode with scalings from the cathode and its support, and with chromium oxide resulting from oxidation of chromium chloride from the bath that was entrained with the product.

In subsequent experiments, a two-piece soapstone cover was used to close the top of the cell, a circular hole to accommodate the cathode support was formed, half in each piece of the cover, and a second hole was formed in one of the cover pieces to permit a constant stream of gas to escape from the cell while substantially cornpletely excluding air therefrom. An additional passage through one of the sections of the soapstone cover was provided for attachment externally to a hydrogen gas supply line for feeding hydrogen gas into the space between the surface of the bath and the soapstone cover. Before charging the pressed sticks of chromium chloride, substantially all air was purged from this space and rcplaced by hydrogen gas from the hydrogen supply line. The chromium chloride was charged through the gas exhaust hole in the soapstone cover and was thus protected from oxidation by the hydrogen atmosphere within the cell as it was melted into the bath. The hydrogen atmosphere was maintained throughout the electrolysis operation, which was otherwise carried on in Substantially the same manner as in the preliminary experiments dcscribed above. Using bath solvent compositions of sodium chloride, straight potassium chloride, and mixavaasos S tures thereof, yields of 70 to 90% of the available chromium were obtained from single batch operations, and the product, after Washing with water and acid, was found to be of excellent purity, ranging from 98.6% to 99% chromium, the impurities consisting principally of small quantities of iron, silicon, and carbon.

With these promising preliminary results, an attempt was made to employ the same process and apparatus on a larger scale, as would be required for any significant commercial production. However, great difficulty was encountered in keeping the chromium deposit on the cathode. The metal kept slipping olf during the run and appeared to be converted in the bath back to chromium chloride, making it impossible to reach an end point so necessary to successful operation, as explained above. Also, for reasons not fully understood, the hydrogen atmosphere above the bath seemed to function less eiiiciently, the recovery of chromium and its purity were low by comparison with the preliminary experiments, and the current eliiciency was very bad.

In view of these problems, I ultimately devised a different kind of cell employing an Inconel pot of the type shown in the accompanying drawing and hereinafter described, the pot being more or less conventional except for the refractory ring about its upper periphery. The pot formed the cathode of the cell in this instance, and the anode was a carbon rod lowered through a central opening in a two-piece soapstone cover of the type also shown in the drawing.

Using this electrolytic cell, approximately 100 pounds f sodium and potassium chlorides in equal parts were charged and melted in the pot, and a hydrogen atmosphere was maintained above the bath while approximately pounds of chromium chloride were charged. The hydrogen atmosphere was maintained throughout the electrolysis in the same manner as in the preliminary experiments, and the bath level was maintained above the lower edge of the refractory ring. Two additional 5 pound quantities of chromium chloride `were added at intervals during the run. Electrolysis was continued with an average potential of about 6 volts applied across the anode and cathode and with an average current of about 800 amperes. After about the theoretically required number of ampere hours of current had been put through the cell the run was terminated, the bath was decanted olf, and the chromium deposit was removed from the pot, washed with water and acid, weighed, and analyzed. Only about 53% of the available chromium was recovered, and the product wascontaminated with an excessive amount of chromium oxide in addition to the small quantities of iron, carbon, silicon, and other impurities found in the products of the smaller scale experiments.

However, when this large scale operation was further modified by introducing hydrogen into the bath in accordance with the present invention, better than 95% of the available chromium was recovered from a single batch operation with the use of only about 86% of the current theoretically required to deposit that amount of metal. The product was 99% pure chromium metal with about 0.57% iron, resulting from iron present as an impurity in the original chromic chloride employed.

When the last described process was carried out in the identical manner and in the same electrolytic cell, except for the use of a similar pot having no refractory ring, the bath became heavily loaded with green chromium oxide in a finely divided state of partial suspension, and prolonging the electrolysis failed to exhaust the bath of chromium chloride. Every effort to recover the bulk of the deposited chromium free from the oxide in the bath was unavailing. The best that could be done was to recover only about 31% of the chromium charged to the cell as chloride, in a relatively pure state, the balance being unrecoverable from the large amount of oxide or remaining unreduced in the bath salts inrspite of pro- 6 longing the electrolysis in a fruitless eli'ort to reach an end point.

From the foregoing, it will be seen that either omitting the forcible diffusion of hydrogen gas through the bath or omitting the use of the refractory ring in the manner described renders the process valueless for any practical purpose. Thus, the combination of both of these features of the invention has been found to be essential.

The apparatus successfully used as described above is illustrated in the accompanying drawing in which Figure l is a somewhat diagrammatic, elevational view of the complete electrolytic cell, shown partly in section, and Figure 2 is a plan view of the cell.

The electrolytic cell, as shown in the drawing, may comprise an Inconel pot 10, preferably having a round bottom. A ring 11 of iron, welded around the upper end of the pot, is shaped to provide an inwardly opening seat or recess into which magnesia or other non-metallic refractory 12, of the character mentioned above, is rammed to form a hard, chemically inert, and electrically non-conductive inner cell surface 13 around the top thereof. The pot 10 may be supported by the ring 11 in any suitable manner, not shown, so that the pot may be tilted to discharge the depleted bath therefrom and" so that heat may be applied to the outside of the pot for melting the initial charge and keeping the bath up to temperature during electrolysis.

Suspended so that it may be lowered into the pot is 'a graphite anode 15 that is also shaped to have a round bottom. The anode is bored longitudinally from its upper end for substantially its full length, and the lower end is drilled to provide a plurality of diverging, downwardly and outwardly directed gas jet passages 16, and, preferably, also a vertical jet passage 17. The upper end of the anode may be -itted with any suitable form of metal plug 18 gripped, together with the upper end of the anode 15, by an electrically conductive cross-arm 19. An electrically conductive cushioning collar 20 of woven copper or the like may, if desired, surround the upper end of the anode to prevent its being broken where gripped by the cross- 'arm 19. Any desired mechanism (not shown) may be employed to mount the cross-arm 19 for vertical movement. The plug 18 is provided with a centrally disposed gas passage 21 terminating at its upper end in `a threaded boss 22 to which a flexible gas line 23 may be attached by means of a fitting 24.

The upper end of the pot is closed by a heat resistant cover of soapstone, or the like, made of two parts 25a and 25h that may be slid laterally in opposite directions to permit free movement of the anode 15 into and out of the bath. A round anode-receiving aperture 26 in the center of the cover is formed, half in the cover section 25a and half in the cover section 25h, and a second 'aperture 27 for rcharging the chromium bearing material is provided in one of the cover sections. To prevent too rapid escape of hydrogen and HCl gases from the aperture 27, a soapstone plate 28 may be laid loosely over the opening. Thus, the only surfaces above the bath which are exposed to gases in the cell are the refractory, upper, inner surface 13, the surfaces of the soapstone cover parts 25a, 25b, and 2S, Iand the surface of the graphite anode 15, all of which are highly resistant to chemical attack by the contents of the pot.

In this case, the positive lead 30 is attached to the cross-arm 19, and the negative lead 31 is attached in any desired manner to the pot 10.

The remarkable improvement in yield, current ediciency, and purity of the product obtained by using the large scale modified lcell shown in the drawing and described above is illustrated by the following example:

Example 1 A bath of lbs. of sodium and potassium chlorides in equal proportions was brought to a temperature of about 850 C. in the pot of the cell shown in the drawing. With hydrogen gas owing through the gas line 23, through the length of the anode, and out the jet openings 116 and 17, the lanode was lowered into the bath until its lower end was only 4 or 5 inches from the bottom of the pot. The soapstone cover sections 25a and 25h were then slid into place with the plate 28 removed, `and `the ow of gas from the anode was continued under these conditions until substantially all air had been purged from the space above the surface of the bath. After passing current through the bath for `a few minutes to assist in removing the last traces of any moisture present in the bath salts, 5 lbs. of -chromic chloride were charged through the aperture 27, and the plate 2S was moved over the aperture. By reason of the loose fit of the plate 2S, gas streamed slowly around its edges from the aperture 27 and was burned, `as indicated in the drawing by the flame 32. The bath level was above the Vcenter of the refractory surface 13 at this time and remained above the lower edge thereof throughout the electrolysis, by reason of the periodic addition of more alkali metal chloride and two further additions of 5 lbs. of chromium chloride each as required to maintain electrolysis.

When a direct current potential of 9.2 volts was applied across the cathode and anode leads 30 and 31 after making the irst addition of chromium chloride, the current rose immediately to 600 amperes. The electrolysis began at 9:0() a. m. and was continued for about 3 hours, the voltage and -current being recorded every few minutes to permit the approximate current consumption to be calculated. The current and voltage remained steady with only very slow changes until the end of the run, as indicated by the following randomly selected readings:

The run was terminated at 1:04, by which time a small amount of a bluish colloidal dispersion had formed in the bath `and was detectable on the surface of a test rod dipped into the bath. Also the bath had changed from an acidic condition to a slightly alkaline condition. The colloidal dispersion was apparently due to liberation of alkali metal after complete depletion of all of the chromium chloride, which accounts for the alkaline condition.

As noted in the above tabulation of data, the approach of the, end point of the run was evidenced by the change of the bath from the characteristic color of the chromium chlorides to a clear, colorless condition, indicating that the bath had been depleted of the chromium salts. The rapid dropping off of the current, with considerable fluctuation, betWeen 1:00 and 1:04 p. m. and the simultaneous appearance of a colloidal dispersion of alkali metal in the bath constituted further unmistakable evidence that substantially all the chromium chlorides had been disassociated and that the run should be terminated.

After withdrawal `of the anode i5 and removal of the cover sections 25a and 2511, the bath was decanted 01T and the chromium deposit, which adhered to the cell wall, was raked out. Analysis of the bath showed only 4a trace of chromium, the amount being so small as to be detectable only qualitatively.

The chromium deposit, after washing and leaching with acid to remove entrained, soluble material from the bath,

was found to constitute better than of the available chromium. Upon analysis, it was found to consist of:

Per `cent Chromium 99.0 Iron 0.75 Carbon 0.14 Balance Undetermined The iron content was due entirely to iron impurities in the chromic chloride (which analyzed 0.25% iron) and may be readily reduced by better control of the purity of this starting material.

Assuming that the chromic chloride was pure, the available chromium from the l5 lbs. of the chloride should have been 4.92 lbs. According to theory, conversion of this quantity of tri-valent chromium to the metallic state should have `required approximately 3450 ampere hours. However, only approximately 2970 ampere hours were consumed, or about 86% of that theoretically required.

From Example 1, it is apparent that hydrogen gas forcibly diffused through the bath itself must perform a chemical reduction of some of the chromium, in addition to preventing reoxidation of the chromium by removing free chlorine as it is released from the anode. Forcing the hydrogen from the lower end of the anode downwardly to the bottom of the pot so that it diffuses upwardly through the entire bath also appears to perform a valuable mechanical function by keeping the bath in a state of mild agitation. This agitation seems to be essential to avoid a tendency for the lower portion of the bath below the anode to remain rich in unconverted chromic chloride and chrornous chloride, thus making possible the complete exhaustion of chromium chloride from the bath. This is important to the recovery of a product free from oxide, for any chromous and chromic chlorides entrained with the product decompose to a considerable extent and form CrzOs, which cannot be removed with any known selective solvent and remains in the product as an impurity.

Thus, it appears that forcibly diffusing hydrogen gas through the bath increases the calculated electrical efliciency to better than 100% by chemical action and makes possible the complete exhaustion of all chromium from the bath, thereby increasing the yield to nearly 100% and improving the purity of the product. As explained above, the refractory ring and maintenance of the bath level above its lower edge are also essential to achieving such results.

The application of the invention to the fused bath electrolysis of other metal chlorides is illustrated in the following additional examples:

Example 2 A bath of 26.4 lbs. of sodium and potassium chlorides in equal proportions` was brought to a temperature of 800 C. in the pot of a cell like that shown in the drawing, but somewhat smaller in size than the one employed in Example l hereof. With hydrogen gas flowing through the gas line 23, through the anode, and through the jet openings 16 and 17 thereof, the anode was lowered into the bath until its lower end was about two or three inches from the bottom of the cell. The Soapstoue cover sections 25a and 25b were then slid into place with the plate 28 removed, and the low of gas from the anode was continued under these conditions until substantially all air had been purged from the space above the surface of the bath, at which time the plate 28 was put in place. With voltage applied, additions of FeClz were made through the aperture 27 at intervals as indicated in the following table, the plate 28 being slid aside for the time required to make the additions. The gases emerging around the plate 28 were burned as before. The bath level was between the upper and lower edges of the refractory surface i3 throughout the run. The current and voltage remained steady during the run of about minutes, ex-

cept for slow changes indicatediby the following selected readings:

Time Amper-es Volts Remarks Added 500 gms. FeClz.

WONOODMOGMMNNUJCB Bath crystal clear.

At 12:48 p. m. the cover was removed, the cathode was withdrawn, and the bath was decanted oi to permit the iron deposit to be raked out. Complete conversion of substantially all of the ferrous chloride was assured by the change in the bath color from the characteristic ferrous chloride yellowish brown color to a colorless crystal clear condition. However, because of the short run, the relatively small amount of material electrolyzed, and the large pot surface from which the deposit had to be recovered, no effort was made to make a complete recovery. The metal actually removed amounted to 523 gms., representing about 78% yield, and would closely approach theoretical with longer runs repeated in the same cell in commercial operations. The recovered metal analyzed 99.9% Fe, with less than 0.01% carbon, and was in the form of sparkling bright crystals of about 40 to 100 mesh size. The current eciency, based on the 100% yield indicated by the exhaustion of the bath, was 72%.

Example 3 A bath of 26.4 lbs. of sodium and potassium chlorides in equal proportions was brought to a temperature of 850 C. in the same apparatus in Example 2. Following the same procedure as in Example 2, but substituting FeCl3 for FeClz, the electrolysis proceeded as indicated by the following data:

Time Amperes Volts Remarks Start 9:58 a. m 100 4. 0 Bath clear. 10:00.." 100 4.0

320 6. 2 }250 gm. FeCls added changing 340 5. 2 bath color to brown. 360 5. 2 360 5. 2 250 gm. FeCla added. 400 5. 2 250 gm. FeCla added. 500 6.0 250 gm. FeClz added. 500 6.0 250 gm. FeCla added. 500 5. 6 250 gm. FeCla added. 500 6.0 250 gm. FeCla added. 500 6.0 250 gm. FeCla added. 500 6. 400 5.0 300 4. 2 Bath color still brown. 300 4' 2 Bath turned progressively lighter 260 4' 2 in color 160 3.8 100 3.8 Bath clear and colorless.

A total of 2 kg. of FeClg was added during the run, this being equivalent to about 690 grams of iron. Of this amount of available iron, 560 gms. or about 81% was recovered in this single batch operation by scraping the metal deposit from the pot rather roughly, no attempt being made to scrape the pot entirely clean. Since the bath was obviously completely depleted of iron chloride and substantially 100% recovery should be obtained from a continuing series of such bath operations in the saine equipment, current efficiency was calculated on an assumed 100 percent Arecovery,tl1e eiciency on this basis being V73 per cent.

Example 4 Again following the procedure of Example 2 and employing the same "type of apparatus, but of a slightly smaller size, a bath of 19.8 lbs. of sodium and potassium chlorides in equal proportions was brought to a temperature of 750 C. Then 1460 gms. of mixed chromic and ferric chlorides in the proportions of 65.5:34.5 were charged into the bath in increments over a period of 11/2 hours with voltage applied across the anode and cathode. The voltage dropped slowly during the run from about 7 volts at the start of the run to about 5 volts at the end, and the current dropped slowly from an initial 400 amperes to a final 220 amperes after about 21/2 hours of operation. The bath color changed from an original light violet to a colorless, clear condition at the end of the run, when the bath was substantially depleted of all iron and chromium chlorides.

About 350 gms. of a sparkling gray ferro-chromium alloy were recovered by roughly scraping the pot at the conclusion of the run after decanting 0E the alkali metal chloride bath. This represented an actual recovery of about S0 per cent of the total available chromium and iron and could have been raised to close to per cent by more careful cleaning of the deposit from the pot. The product analyzed better than 99 per cent ferro-chromium and only about 0.03 per cent carbon. The ratio of chromium to iron in the product was substantially the same as the ratio of available chromium to available iron in the mixed chloride charge material, namely, 189:1. The very low carbon content of the product is particularly noteworthy.

Example 5 Following the same procedure as in Example 4 and employing the same type of apparatus, a bath of 19.8 lbs. of sodium and potassium chlorides in equal proportions was brought to a temperature of 850-900 C. Then 1500 gms.'of MnClz were charged into the bath in increments over a period of about an hour with voltage applied across the anode and cathode. The voltage ranged from about 7 volts to about 5 volts and the current from about 600 amperes to about 350 amperes during the run of about 5 hours and 20 minutes. The bath color Ygradually changed from pink to a clear colorless condition toward the end of the run, the color substantially disappearing before the bath was completely depleted of manganese chloride, and operation was continued until the desired end point was reached. Of the 654 gms. of available manganese, 600 gms. was recovered by roughly scraping the deposit from the pot at the conclusion of the run, this crude recovery being 91.74 per cent of the available manganese. The product analyzed 99.35 per cent manganese and only 0.08 per cent carbon.

To make fully effective use of the hydrogen introduced linto the bath in accordance with the present invention, and also to permit recovery of the hydrochloric acid gas formed in the cell and the excess hydrogen employed, the cover for the cell is essential. The cover also prevents any loss of the electrolyte which might be caused by splashing of the bath due to agitation of the surface thereof. in addition, the resulting protective atmosphere maintained over the surface of the bath prevents any possible oxidation of the bath material that may be splashed upwardly during the process, prevents entrapment of atmospheric oxygen by such splashing of the bath surface, and protects the charge material from oxidation while it is being introduced into the cell and melted into the bath.

The opening 27 shown in the drawing may be as large as needed for introducing the chloride to be electrolyzed through the protective atmosphere, the cover plate 28 providing suicient resistance to escape of gases from the cell to maintain a gas pressure of a small fraction of one atmosphere (gauge) within the cell and prevent infiltration of air. If desired, any suitable gas recovery system (not shown) may be connected to this opening 27 while the cell is in operation; or an additional opening (not shown) in the cover for permanent connection to a gas recovery system may be provided.

As indicated above, the process of the present invention is particularly suited for the recovery of polyvalent metals from their chlorides, whether the chlorides employed are the higher or lower chlorides of such metals. Apparently, the chemical and/ or thermic reduction of such chlorides by hydrogen during the process is most pronounced reducing higher chlorides of polyvalent metals to lower chlorides or to elemental form. This seems to involve the partial or complete reduction of higher chlorides that may be charged into the cell as a source of the metal or metals to be recovered and also the partial or complete reduction of .higher chlorides formed in the bath from initially charged lower chlorides by the attack of chlorine released at the anode before it is taken up by the hydrogen. It also appears that some of the metal to be recovered may be reduced to its elemental form entirely by the chemical and/ or thermic action of hydrogen.

While the invention has been illustrated herein by examples employing specific metal chlorides as the source of the metals deposited on the cathode, it will be understood that the invention is not limited to the electrolysis of chlorides of these specific metals. Numerous variations of the apparatus and the details of the process may also be employed without departing from the true spirit and scope of the invention, as will be apparent to those skilled in the art. It will also be understood that the particular form of apparatus shown in the drawingV is suitable for use in other electrolysis operations in which a gas is to be forcibly introduced into the hath and dispersed or diffused therethrough.

This application is a continuation-in-part of my prior application Serial No. 214,988, tiled March 10, 1951, which was a continuation-in-part of my still earlier application Serial No. 201,089, filed December 16, 1950, both nowbeing abandoned.

Having described my invention, I claim:

l. The process of depositing metals in solid form from a fused salt bath in an electrolytic cell, comprising employing a closed, heat resistant, metal alloy pot as the cathode of said cell, the upper portion of the side wall of the pot about its entire periphery being constructed to provide an upper inner surface of non-metallic refractory material that is substantially inert to chemical attack by the contents of the pot, electrolyzing in said pot a fused salt bath consisting essentially of alkali metal chloride and the salt of at least one metal to be deposited, the salt being selected from the class consisting of chromium chloride, iron chloride, manganese chloride, and mixtures thereof, carrying out the electrolysis in the presence of hydrogen gas diffused through the fused salt bath, and, continuously during electrolysis, maintaining the surface of the fused salt bath above the lower edge of said upper inner surface of the pot.

2. The process of claim 1 in which electrolysis is continued until the bath is substantially completely depleted of chloride of metal being deposited. i

3. The process of claim l in which electrolysis is continued until the bath becomes alkaline as a result of liberation of alkali metal from the alkali metal chloride in the bath.

4. The process of claim l in which the metal being deposited is iron.

5. The process of claim 1 in which the metal being deposited is manganese.

6. The process of claim 1 in which the metal being deposited is chromium.

7. The process of claim l in which said fused salt bath contains chlorides of iron and chromium, whereby said iron and chromium are deposited together as a ferrochromium alloy.

8. The process of depositing metals in solid form from a fused salt 'oath in an electrolytic cell, comprising employing a closed, heat resistant, metal alloy pot as the cathode of said cell, the upper portion of the side wall of the pot about its entire periphery being constructed to provide an upper inner surface of non-metallic refractory material that is substantially inert to chemical attack by the contents of the pot and is electrically substantially non-conductive, electrolyzing in said pot a fused salt bath consisting essentially of alkali metal chloride and the salt of at least one metal to be deposited, the salt being selected from the class consisting of chromium chloride, iron chloride, manganese chloride, and mixtures thereof, carrying out the electrolysis in the presence of hydrogen gas diffused through the fused salt bath, and, continuously during electrolysis, maintaining the surface of the fused salt bath above the lower edge of said upper inner surface of the pot.

9. The process of claim 8 in which electrolysis is continued until the bath is substantially completely depleted of chloride of metal being deposited.

10. The process of claim 8 in which electrolysis is continued until the bath becomes alkaline as a result of liberation of alkali metal from the alkali metal chloride in the bath.

11. The process of claim 8 in which the metal being deposited is iron.

l2. The process of claim 8 in which the metal being deposited is manganese.

13. The process of claim 8 in which the metal being deposited is chromium.

14. The process of claim 8 in which said fused salt bath contains chlorides of iron and chromium, whereby said iron and chromium are deposited together as a ferrochrornium alloy.

15. The process of claim 8 in which said hydrogen gas is introduced directly into the fused salt bath below the surface thereof.

16. The process of claim 8 in which said hydrogen gas is introduced directly into the fused salt bath below the surface and adjacent the bottom thereof.

17. The process of claim 8 in which said hydrogen gas is introduced directly into the fused salt bath below the surface and adjacent the bottom thereof and at least partially in a downward direction.

18. An electrolytic cell comprising a metallic pot having a ring of a non-metallic refractory material, that is substantially inert to chemical attack by chlorine and hydrogen chloride at elevated temperatures, extending entirely around the periphery of the upper end of the pot substantially as a continuation of the side wall thereof to form an upper, inner, refractory, pot surface about said periphery, a cover of such refractory material resting on said refractory ring for closing the pot and forming a continuation of said upper, inner, refractory, pot surface over the upper end of the pot, said cover having an anode receiving opening therethrough and an elongated graphite electrode having a longitudinal passage extending therethrough from the upper end to adjacent its lower end, at least one generally downwardly directed discharge orifice in said anode adjacent the lower end thereof for discharging gas from said passage, means supporting said anode for vertical movement into and out of said cell through said anode-receiving opening, and means for feeding a stream of gas into the upper end of said anode for travel downwardly through said passage and discharge orifice.

References Cited in the le of this patent UNITED STATES PATENTS 466,460 Edison Ian. 5, 1892 1,845,266 Griswold Feb. 16, 1932 2,419,383 Ames Apr. 22, 1947 FOREIGN PATENTS 709,742 Germany Aug. 26, 1941 OTHER REFERENCES Transactions of the Electrochemical Society, vol. 87 (1945.), pages 551 thru 567 of article by Kroll.

Claims

1. THE PROCESS OF DEPOSITING METALS IN SOLID FORM FROM A FUSED SALT BATH IN AN ELECTRTOLYTIC CELL, COMPRISING EMPLOYING A CLOSED, HEAT RESISTANT, METAL ALLOY POT AS THE CATHODE OF SAID CELL, THE UPPER PORTION OF THE SIDE WALL OF THE POT ABOUT ITS ENTIRE PERIPHERY BEING CONSTRUCTED TO PROVIDE AN UPPER INNER SURFACE OF NON-METALLIC REFRACTORY MATERIAL THAT IS SUBSTANTIALLY INERT TO CHEMICAL ATTACK BY THE CONTENTS OF THE POT, ELECTROLYZING IN SAID POT A FUSED SALT BATH CONSISTING ESSENTIALLY OF ALKALI METAL CHLORIDE AND THE SALT OF AT LEAST ONE METAL TO BE DEPOSITED, THE SALT BEING SELECTED FROM THE CLASS CONSISTING OF CHROMIUM CHLORIDE, IRON CHLORIDE, MANGANESE CHLORIDE, AND MIXTURES THEREOF, CARRYING OUT THE ELECTROLYSIS IN THE PRESENCE OF HYDROGEN GAS DIFFUSED THROUGH THE FUSED SALT BATH AND, CONTINUOUSLY DURING ELECTROLYSIS, MAINTAINING THE SURFACE OF THE FUSED SALT BATH ABOVE THE LOWER EDGE OF SAID UPPER INNER SURFACE OF THE POT.