US2677596A - Process for separating ferrous chloride from a mixture of fecl2 and cocl2 - Google Patents

Description

Patented May 4, 1954 UNITED OFFICE PROCESS FOR SEPARATING FERROUS CHLORIDE FROM [1 MIXTURE OF FeClz AND C0012 N0 Drawing. Application November 15, 1952, Serial No. 320,835

The present invention relates to a process for recovering the cobalt values from a gaseous mixture of FeClz and C'oCh, and more particularly for treating such a gaseous mixture so as to convert any cobalt chloride present therein to a nonvolatile form of cobalt, such as metallic cobalt and without appreciably altering the composition of the ferrous chloride originally present.

Briefly, this is done by reacting any cobalt chlo- 6 Claims.

ride originally present in the gaseous mixture 1' with metallic iron to form ferrous chloride and elemental cobalt. The ferrous chloride thus formed is passed to a desired point, along with the ferrous chloride originally present, leaving he cobalt in some metallic form. It is recognized that the metallic cobalt formed may not remain separate as such, but may possibly combine with some metallic iron present and/or that which is provided for reaction purposes, as aforesaid, to form some cobalt-iron alloy which is useable commercially. However, the broad purpose is socured, which is to convert the cobalt to non-volatile and elemental or metallic form, the ferrous chloride produced and that originally present being separated for such use thereof as may be desired.

The present application is one of a group of seven co-pending applications, all the inventions of the same inventors, four of these cases relating to nickel and reactions involving nickel and the other three relating to cobalt and reactions involving cobalt. The cases may be compared and distinguished as follows:

a. Ser. No. 160,821, filed May 8, 1950, relates to treating a starting material in the non-gaseous 1 (liquid and/ or solid) form and consisting essentially oi NiClz and FeClz, and wherein there is more FeClz than NiClz. This starting material is tr ted i a reaction or sublimation zone, in the e of a certain amount of metallic iron, to produce a product which is vaporized FeClz, removed the sublimation zone as such, leaving only metallic iron and metallic nickel in the sublimation zone.

2;. No. 210,433, filed February 10, 1951, relates to treating a starting material, essentially the same as Ser. No. 160,821, except that in this instance the starting material is introduced into the reaction zone in a solely gaseous state. The reaction and the final products are substantially the same as in Ser. No. 160,821.

0. Ser. No. 160,822, filed May 8,1950, relates to the preventing or minimizing of a possible back reaction. The starting material consists essentially of ferrous chloride and metallic nickel. It

is desired to separate these materials without permitting, or at least minimizing, a possible back reaction therebetween which would produce nickel chloride and metallic iron. To do this, the starting material is introduced into a reaction or sub.- limation zone in which metallic iron is present. The starting material as introduced into this zone is in a non-gaseous state. The final products are substantially pure ferrous chloride vapor, which is removed from the Zone as such, and metallic nickel and metallic iron which remain in the zone.

01. Ser. No. 294,058, filed June 17, 1952, discloses a process for treating a startin material substantially the same as in Ser. No. 160,821. This case has been amended so as to preclude the claims reading upon separation of the products of the reaction by vaporization as in Ser. No. 150,821; but in this case the separation following the reaction is effected by withdrawing the metallic nickel produced (by a reaction the same as in Ser. No. 160,321) by withdrawing massive pieces of iron, to which the depqsited nickel adheres physically, from the fused bath of the starting materials. Neither the ferrous chloride nor any nickel chloride present is intentionally volatilized. This application is junior to application Ser. No. 160,821, so that the only subject matter claimable therein is subject matter which cannot be supported by the disclosure of Ser. No. 160,821; all the common subject matter being claimed in Ser. No. 160,821.

e. Ser. No. 291,816, filed June 4, 1952, is the first of the three cobalt applications and is the cobalt counterpart of the nickel case, Ser. No. 160,821, distinguishing therefrom by being directed to cobolt and its chloride, rather than nickel and its respective chloride. This case is further distinguished in the degree of purity of the sublimate.

f. Ser. No. 320,835, filed November 15, 1.952 (the present application), is the cobalt counterpart of nickel case, Ser. No. 210,433. It distinguishes from cobalt case Ser. No. 291,816 in the same way discussed above as to the respectively corresponding nickel cases, while distinguishing from its corresponding nickel case, Ser. No. 210,433, by the differences between cobalt and nickel.

9. Ser. No. 321,514, filed November 19, 1952, is the cobalt case corresponding to the nickel case Ser. No. 160,822. It distinguishes from the other cobalt cases as discussed above in respect to the respectively corresponding nickel cases, while distinguishing from its nickel counterpart, Ser. 160,822, by the diiierences between nickel and cobalt.

In the prior art illustrated, for example, in the U. S. patents to Kroll Nos. 2,396,792-3-4, all issued March 19, 1946, an attempt was made to separate ferrous impurities from nickel. In accordance with these Kroll patents it was at tempted to carry out a transfer reaction between iron in elemental form and nickel chloride, so as to form nickel in elemental form and ferrous chloride. This was accomplished by Kroll, in

accordance with one of the patents aforesaid, in A a fused salt bath. In another of the Kroll patents aforesaid, the same reaction was attempted to be accomplished without the bath, but conditions otherwise being substantially the same. In the third of these patents, the purpose was to chloridize nickel and iron in a fused bath by passing chlorine therethrough; and the reaction aforesaid was relied upon to get what is in its final result a selective chloridizing of the iron without chloridizing the nickel. In all the Kroll disclosures the purpose was basically to provide pure nickel and to free this nickel from iron, among other metals, as an impurity, the iron being present in a very small amount, in respect to the amount of nickel present. attempt, in accordance with the Kroll disclosure, to effect a physical separation simultaneously with the principal chemical transfer reaction, as aforesaid, but the physical separation was to be effected subsequent to the chemical reacis a reversible type reaction which proceeds only I):

to some intermediate equilibrium stage. i order to bias the equilibrium in a desired direction, Kroll resorted to the expedient of adding a substantially large excess of nickel chloride, 25% to 50%, for example.

In accordance with the present invention, the basic material to be worked upon in the present process contains cobalt, rather than nickel, and further is contemplated to contain a relatively large percentage of iron, principally in the form of ferrous chloride, as compared with the amount of cobalt chloride. For example, in typical ores containing iron and cobalt, there is contained more than 50 times as much iron as cobalt.

This difierence in raw material however, is not presently relied upon as a basic distinction from the prior art, but is mentioned as illustrating a difference in the type of material which the present invention is adapted to treat as compared with the material contemplated for use in the prior art.

In accordance with the present invention, it is contemplated that there will usually be present, during the reaction, suflicient iron in elemental form to react with all the chlorine initially introduced in the form of cobalt chloride, so as to form therefrom ferrous chloride by the chemical transfer reaction aforesaid. Preferably, also, there will be a reasonable excess of elemental iron initially present over such equivalent or stoichiometric proportions as hereinafter more particularly set forth. The reaction is basically the same as that which was contemplated, in accordance with the prior art teaching as described, except that in the prior art the reaction was between metallic iron and nickel chloride, whereas in the present invention the reaction is between metallic iron and cobalt chloride. However, in the present case, the reaction is caused to carry through to substan- There was no tial completion, irrespective of the reversible character of the reaction and some intermediate equilibrium end point, by separating from the process the ferrous chloride substantially as it is formed.

The purpose of the present invention in its preferred form is two-fold: first, to convert substantially all the cobalt chloride present in the gaseous phase in the starting material to cobalt in metallic form, which may be used as such in some way known to the art; and second, to separate cobalt in chloride and/or metallic form from ferrous chloride, so that the latter may be diverted to a suitable point at which it may be used or converted into some other useful product, the details of such use or conversion forming no part of the present invention.

Summarizing the present invention, therefore, it comprises establishing in a reaction zone a body of metallic iron, which if in solid form,

is preferably in a mass which is gas-pervious, such, for example, as a mass of steel wool, and which preferably, also, is in such a form as to present a relatively large surface area per unit of weight. This provides a large surface of solid iron for reaction with the gas. Alternatively, if the iron is in liquid form, the gaseous mixture may be bubbled therethrough or otherwise passed in contact therewith, this gaseous mixture preferably being supplied to the bath of molten iron at a point below the surface thereof. This iron may then react with the cobalt chloride in the gas, as hereinafter specifically set forth, the cobalt produced going into solution in the metallic iron (assuming molten iron) and the ferrous chloride being released on the surface of the bath as substantially pure ferrous chloride. The use of a bath of molten iron lends itself particularly to a combined process, including as one step, the present process and, as a second step, a process for the further treatment of ferrous chloride, which, per se, forms no part of the present invention and which, therefore, is not more particularly disclosed herein.

The gaseous mixture useful as the input material for the present process must contain ferrous chloride and cobalt chloride, but may also contain some one or more gases which are inert insofar as the contemplated reaction is concerned, for example, nitrogen. By the use of some additional, relatively inert gas, as aforesaid, it is possible to use a gaseous mixture con-- taining less than 100% of both ferrous chloride and cobalt chloride respectively and in total, so as to permit the operation of the process at temperatures below the boiling points of both ferrous chloride and cobalt chloride.

The preferred temperature range contemplated for use in the present case is from apt/"t 800 *1. up to the temperature at which the metallic balt produced by the reaction would have a substantial vapor pressure. More particularly, the preferred temperature range is from about 990' C. up to the melting point of iron for the phase of the process in which the iron present in the reaction zone is solid, while temperatures necessary for the maintenance of iron in the liquid phase may be used when operating in accordance with that phase of the invention.

Turning now to the details of the present invention, the first factor to be considered is the composition of the gaseous mixture being passed into the reaction zone and the origin thereof. The particular manner in which the mixtures of ferrous chloride and cobalt chlorides are preany acrea e pared for treatment, in accordance with the present process is no part of the present invention. However, the process is: applicable in treating some ores which contairr both and cobalt with a very high ratio of iron to cobalt. If such an ore be chloridized and the chlorides of iron and cobalt vaporized from the chloridized ore, a mixture in substantial accordance with that contemplated in the present invention, is produced. Similarly, in recovering. iron and cobalt from alloy scrap containing both these metals, one method of recovering such metal values is by chloridizing the entire scrap, then treating the volatilized chlorides in accordance with the present invention.

The manner in which the chlorides of iron and cobalt are volatilized is also no part of the present invention, it being contemplated generally that sufficient heat will be provided so thatthe vapor pressure of these two chlorides will be sufficient in conjunction with thepartial pressure of inert gas present to cause volatilization of ferrous and cobalt chlorides to the extent necessary or desired, in accordance with the present invention.

As to composition, it is, of course, contemplated that the gaseous mixture being treated may consist solely of FeClz and COClz vapors. On the other hand, it is often desired, in accordance with the present invention, to operate at temperatures below the boiling points of these chlorides. In order to do so, some relatively inert gas may be admixed with the vapors of these two chlorides, for example, nitrogen, or even other chloride vapors. When operating at temperatures above the boiling points of both FeClz and C0012, some inert gases may be admixed therewith. It is further contemplated that any gas or gases, which are inert insofar as the present reaction is concerned, may be used along with the mixed chloride vapors, nitrogen being one example which is operative underthe circumstances.

The reaction may take place in any type of apparatus useable for this general purpose, particular details of the apparatus forming. per se, no part of the present invention. For this reason, the place or apparatusirr which the reaction is to be carried on, is herein termed a reaction zone. It is contemplated that any solid-to-gas or liquid-to-gas apparatus known to the art, or which will occur to those skilled in the art from the present description and which is or may be made appropriate to the carrying on of the present process, may be employed.

The iron, which must-be present in the reaction zone in accordance with the present invention, is particularly described as a body of metallic iron. It is contemplatedthat this may be a solid metallic body, such as a bolt or rod, or that it may be solid material in anyd'esired' form. Preferably, however, the metallic iron is present in a form providing a maximum surface area for contact with the gas in respect to the amount of iron present. Thus a relatively finely divided form of iron is usually preferable. Specifically, for example, it is contemplated, and it has been found, that steel wool is efficacious for this purpose. It is further contemplated that some form of sponge iron may be used in a mass which is preferably gas-pervious and which provides a very large surface per unit of weight. On the other hand, if sumcient surface is provided, on which the reaction may take place, the particular form of the iron present is not critical.

It is also specifically contemplated, in accord ance with the present invention, that the iron may be a molten bodyor bath and that the gaseous mixture may be bubbled therethrough or otherwise brought into contact therewith as by being supplied to the bath at a point or points below the surface thereof. Here again, the arrangement must be such that there isafforded adequate time for the chemical reaction desired to take place.

The chemical reaction question, which is generally known, may be expressed. by the equation:

CoCl2+Fe=Co+FeClz This reaction has been found to be reversible in character, as set forth above. In order that the reaction shall proceed from left to right, as hereinabove given in the equation, the FeCh produced is removed from the vicinity of the reaction, i. e., from the reaction zone, to a suitable delivery point. It is also practically necessary that adequate surface area of the iron be provided, so that even if some cobalt were to coat some of the iron surface originally provided, there'would be adequate free surface area of iron available for contact and for reaction with the remaining gases.

If substantially all the cobalt is tobe removed from the 'ases, the iron must be present in an amount at least equal to the amount required to react with all the chlorine introduced in the gases to be treated as CDC-1a to form FeClz. In the usual course, however, there willbe several times this much iron present as it is, of course, possible for some of the iron to be chemically shielded or blocked by the chemical reaction, which might prevent reaction of the so-shielded iron with remaining gases or gases following the first gases. For this reason, it is usual in practicing the present invention, to provide a considerable amount of iron in excess of the strict stoichiometric requirements. The particular degree of excess of iron is not critical as is more particularly pointed out in the subsequent examples, but the purpose of the excess is as aforesaid, to insure the maximum completion of the reaction.

The temperatures in the reaction zone have been stated particularly to be within the limits of about 800 C. up to the temperature at which metallic cobalt has a substantial vapor pressure. This broad range may be subdivided into two ranges, with the dividing line therebetween set at the melting point of the iron present. These two ranges and the limits thereof, will now be considered.

The boiling points of 00012 and FeClz at atinospheric pressure are about 1049 C. and 1027 C. respectively. Therefore, at temperatures below this general boiling point range, one or both of these materials will be below its boiling point. The present process is operable, however, below this boiling point range by having an inert gas present to provide a partial pressure forming a part of the total pressure of the gaseous mixture, the remainder being made up of the partial pressures of FeClz and C0012 respectively.

It has been found by experiment, as hereinafter set forth in the particular examples which follow, that when the process is attempted to be oper ated at temperatures substantially below about 800 C. the vapor pressure of the mixtures of CoCh and FeClz to be" worked upon are so low that substantial amounts of vapor cannot be provided for contact with the metallic iron, Within a reasonable time. For this reason, therefore, the temperature of about 800 C. is stated as the low limit of temperature, in accordance with the present invention. This is so, not because the temperature of 800 C. is narrowly critical, but because it represents a fair approximation of the economic low limit of the process.

From many points of view, it is desirable to operate with the iron in the reaction zone in solid form. This enables the process to be carried out using relatively simple apparatus, as compared with apparatus which would be required if the iron in the reaction zone were to be in liquid form. As such, there is a top limit for the temperature in the reaction zone at the melting point of iron for one preferred phase of the process. From the point of view of the present process alone, this lower range is usually preferred.

Considering, however, the joint process above referred to, wherein the present process is but the first step and a process for treating ferrous chloride vapors in a further way (not herein disclosed and forming no part of the present invention) is a second step, it may be desirable to use the iron in the reaction zone in the liquid state. For this reason, therefore, the present invention is intended to include temperatures between the melting point of this iron and the temperature at which the metallic cobalt formed by chemical reaction in the bath would have a substantial vapor pressure. The reason for setting this upper limit is that the metallic cobalt is to be recovered in non-volatile form and it is not desired to operate at temperatures so high that substantial amounts of metallic cobalt vapor may be former. which would escape with the ferrous chloride vapor.

The process of the present invention is further illustrated by the following particular examples:

Example I In order to illustrate the necessity for maintaining the temperature in the reaction zone at at l ast 800 C., a comparison may be made between the following experimental tests:

In the first of these tests, a solid material consisting of about 30 grams of a mixture of C0812 and FeCiz, containing 97.5% E3012 and 2.5% C0012. by weight, was heated to 700 C. and a current of nitrogen passed over the solid material. In four hours, only about 0.5 gram of this solid material had been vaporized, indicating that at this temperature the vapor pressures of F6012 and COClZ were insumcient to provide enough vapor to carry on the desired vapor phase reaction on an economic basis.

In the second test, a solid material consisting of a mixture of CoClz and Feclz having the same composition as above described was heated to 800 C. and a current of nitrogen passed over this solid material to sweep the vapors evolved therefrom through a plug of steel wool, which was also maintained at 800 C. In three hours, it was found that about 40% of the FeClz-COCl-z mixture had been vaporized and the vapor passed through the plug of steel wool. The iron content of the steel wool plug was such that it amounted to 40.4 times the amount of metallic iron theoretically necessary to react with all the chlorine in the COClz that passed through the plug. At the end of the reaction period, it was found that 96.4% of the COClz that had been vaporized had been converted to an insoluble form and remained as metallic cobalt deposited on the steel wool of the plug. The gases after passing through the plug consisted essentially of substantially pure FeClz plus nitrogen.

It may thus be concluded that the minimum temperature at which the vapor pressures of FeClz and C0012 are sumcient for the process of the invention to be carried on at a reasonable rate is at least about 800 C.

Example II Although the temperature of 800 C. set forth above is the practical minimum at which the process may be successfully operated, it is found that there are certain advantages in operating at temperatures somewhat above this minimum. When the mixture of ferrous chloride and cobalt chloride to be treated is initially in the solid form, it is found that it is possible to vaporize the mixture in a much shorter time if a higher temperature is employed.

When a mixture of ferrous chloride and cobalt chloride, containing these materials in the same proportions as employed in tests given in Example I, was heated to a temperature of 900 C. in a current of nitrogen, it was found that substantially all of the mixture was completely vaporized in about two hours. When the vapors produced in this way were contacted with a steei wool plug, also at a temperature of 900 C. it was found that 96.9% of the cobalt chloride in the vapor mixture had been converted to metallic cobalt which was retained on the steel wool plug. The remaining gases beyond the plug consisted essentially of substantially pure ferrous chloride plus the nitrogen which was used as aforesaid. The iron contained in the steel wool plug in this case amounted to 8. 31 times the amount of iron theoretically necessary to react with all the chlorine in the cobalt chloride which passed through the steel wool (iron) plug.

When the process was operated at the temperature of 1000 C. utilizing the same proportions of FeClz and C0C12, it was found possible to vaporize substantially all of the material in the starting mixture in thirty minutes. About 89.5% of the COClz vapor was converted to metallic cobalt by contact with the steel wool plug, which in this case contained an amount of metallic iron 9.22 times the amount of iron theoretically necessary to react with all the chlorine in the COClz passed in contact therewith.

Example III Although mixtures of ferrous and cobalt chlorides will be most often found to contain a considerable preponderance of ferrous chloride, and the examples given above have illustrated the application of this process to such commonly occurring mixtures, an eifective separation as aforesaid, can be obtained using mixtures containing substantially higher percentages of C0C12.

For example, a mixture consisting of of COClz and 20% of FeClz by weight was vaporized in a current of nitrogen and the resulting gaseous mixture was contacted with 4.5 grams of steel wool (about 1.63 times the stoichiometric amount) all the reactants being maintained at a temperature of about 1000 C. After all the vapor had passed through and in contact with the steel wool, it was found that about 86.6% of the C0Cl2 had been converted to metallic cobalt, which deposited on the steel wool plug. Thus, the effective recovery of cobalt from the mixture of these chlorides may be obtained even when relatively high percentages of cobalt are present in the original starting mixture.

Erample IV Although it is generally desirable to provide sufficient metallic iron to amount to a substantial excess over that theoretically necessary to react with all. of the cobalt chloride vapor contacted therewith, it has been found that this excess need not be particularly large.

Thus, when a mixture of ferrous chloride and cobalt chloride was contacted at a temperature of 1000 C. with a plug of metallic iron, equal to only 1.74 times the amount of iron theoretically necessary to react with all of the chlo rine in the cobalt chloride contacted therewith,

90.1% of the cobalt chloride was converted to metallic cobalt and deposited on the steel wool plug. This may be compared with the results obtained under somewhat similar conditions where a steel wool plug containing sufficient metallic iron to equal 9.22 times the amount of iron theoretically necessary to react with all the cobalt chloride brought in contact therewith. In this latter case 89.5% of the cobalt chloride was converted to metallic cobalt.

While there has been disclosed herein a principal process certain specific variants thereof, other alternatives and equivalents will occur to those skilled in the art from the foregoing disclosure. We do not wish to be limited, therefore, except by the scope of the appended claims, which are to be construed validly, as broadly as the state of the prior art permits.

What is claimed is:

l. The process of separating ferrous chloride from a solely gaseous mixture containing a substantial proportion of C0012 and an amount of FeClz at least as great as the amount of said COC12, comprising the steps of establishing in a reaction zone a body of metallic iron, passing said solely gaseous mixture into and through said reaction zone so as to contact it with said metallic iron during its passage therethrough and so as to provide a contact time between said solely gaseous mixture and said metallic iron in said zone adequate for chemical reaction therebetween for a conversion of all of the cobalt chloride present to a non-volatile form of cobalt, said body of metallic iron being stolchiometrically sufiicient in amount to react with all of the chlorine combined with cobalt as COClz in said gaseous mixture which is introduced into and passed through said reaction zone in contact with said metallic iron so as to form FeClz, maintaining the temperatures of said gaseous mixture and of said metallic iron in said reaction zone in the range of about 800 C. up to the boiling point of the metallic constituents present, and separating ferrous chloride, which is substantially free of CoClz, by withdrawing ferrous chloride vapor from said zone to provide substantially pure separated ferrous chloride as one product of the process.

2. The process in accordance with claim 1, wherein said body of metallic iron consists essentially of a body of steel wool.

3. The process in accordance with claim 1, wherein said body of metallic iron is retained in the solid state, and wherein said temperature range is from about 800 C. up to the melting point of the metallic constituents present.

l. The process in accordance with claim 1, wherein said gaseous mixture includes at least one additional gas which is inert in the process, whereby to enable the process to be operated below the boiling points of both CoClz and FeClz.

5. The process in accordance with claim 1, wherein said body of metallic iron is a body of molten iron, wherein the step of passing said gaseous mixture through said reaction zone is effected by passing said gaseous mixture in intimate contact with said molten iron, and wherein said temperature range is that in which said metallic iron will be in the molten state.

6. The process in accordance with claim 1, wherein said body of metallic iron is a body of molten iron, wherein the step of passing said gaseous mixture through said reaction zone is effected by passing said gaseous mixture into said molten iron from at least one point below the surface thereof, so that said gaseous mixture bubbles up through said molten iron, and wherein said temperature range is that in which said metallic iron will be in the molten state.

References Cited in the file of this patent UNITED STATES PATENTS Number Name Date 2,396,792 Kroll Mar. 19, 1946 OTHER REFERENCES J. W. Mellor: Modern Inorganic Chem,

pages 432, 433, new impression of eighth ed., January 1945. Longmans, Green and C0., N. Y.

Lewis: An Outline of First Year College Chem. (1945 ed), pages 48-50. Barnes and Noble, Inc., N. Y.

Claims (1)

1. THE PROCESS OF SEPARATING FERROUS CHLORIDE FROM A SOLELY GASEOUS MIXTURE CONTAINING A SUBSTANTIAL PROPORTION OF COCL2 AND AN AMOUNT OF ECL2 AT LEAST AS GREAT AS THE AMOUNT OF SAID COCL2, COMPRISING THE STEPS OF ESTABLISHING IN A REACTION ZONE A BODY OF METALLIC IRON, PASSING SAID SOLELY GASEOUS MIXTURE INTO AND THROUGH SAID REACTION ZONE SO AS TO CONTACT IT WITH SAID METALLIC IRON DURING ITS PASSAGE THERETHROUGH AND SO AS TO PROVIDE A CONTACT TIME BETWEEN SAID SOLELY GASEOUS MIXTURE AND SAID METALLIC IRON IN SAID ZONE ADEQUATE FOR CHEMICAL REACTION THEREBETWEEN FOR A CONVERSION OF ALL OF THE COBALT CHLORIDE PRESENT TO A NON-VOLATILE FORM OF COBALT, SAID BODY OF METALLIC IRON BEING STOICHIOMETRICALLY SUFFICIENT IN AMOUNT TO REACT WITH ALL OF THE CHLORINE COMBINED WITH COBALT AS COCL2 IN SAID GASEOUS MIXTURE WHICH IS INTRODUCED INTO AND PASSED THROUGH SAID REACTION ZONE IN CONTACT WITH SAID METALLIC IRON SO AS TO FORM FECL2, MAINTAINING THE TEMPERATURES OF SAID GASEOUS MIXTURE AND OF SAID METALLIC IRON IN SAID REACTION ZONE IN THE RANGE OF ABOUT 800* C. UP TO THE BOILING POINT OF THE METALLIC CONSTITUENTS PRESENT, AND SEPARATING FERROUS CHLORIDE, WHICH IS SUBSTANTIALLY FREE OF COCL2, BY WITHDRAWING FERROUS CHLORIDE VAPOR FROM SAID ZONE TO PROVIDE SUBSTANTIALLY PURE SEPARATED FERROUS CHLORIDE AS ONE PRODUCT OF THE PROCESS.