US2822258A

US2822258A - Method of producing metals from their chlorides

Info

Publication number: US2822258A
Application number: US486591A
Authority: US
Inventors: Jordan James Fernando
Original assignee: Individual
Current assignee: Individual
Priority date: 1955-02-07
Filing date: 1955-02-07
Publication date: 1958-02-04
Anticipated expiration: 1975-02-04

Description

J.. F. JORDAN Feb, 4, 1958 METHOD oF PRODUCING METALS FROM THEIR cHLoRIDEs s sheets-sheet 1` Filed Feb.- 7, 1955 FIG.r l

' Feb. 4, 19,58 J. F. JORDAN METHOD oF PRoDucING METALS AFRQM THEIR cHLoRIDEs 3 Sheets-Sheet 2 IN V EN TOR:

Feb. 4, 1958 v J. F. JORDAN METHOD OF PRODUCING METALS FROM THEIR CHLORIDES 3 sheets-sheet 5l Filed Feb. '7, 1955 & M5323 'products.

chloride containing iron particles is caused to ow into v feasible or safe to pour water (B. P. 373 K.) on the surface of a pool of molten lead (M. P. 601 K.).

Reactions of the subject type are strongly exothermic. A number of investigators, including the present inventor, have suggested that advantage be taken of this reaction heat to the end that the products issuing from the reaction zone are at a temperature that lies above the melting point of the product metal, thereby causing the byproduct reducing agent chloride `to be in gas phase so that it will in no way interfere with the deposition of the molten product metal particles which are caused to impinge into contact with the surface of the pool of molten product metal in the collecting phase of the process.

The main di'lculty with this proposal is that the heat released by these reactions is not sufficient to raise the reaction products to the desired temperature without preheating the reactants sufficiently to lift them over 'Y their respective boiling points, with the result that, in

order to gain the desired temperature level, the reactants must be introduced into the reaction zone in gas phase. The resulting vapor phase reaction yields product metal particles whose state of subdivision is so ne that it resembles smoke. While product metal particles can be successfully deposited by such a hot reaction product stream, it is desirable that the eiiciency of the deposition be substantially increased, so as to increase the process yield. The efciency of the deposition is increased as the fneness of the product metal particles is decreased, the best results being obtained when the product metal particles are produced in a reaction in which the reaction zone is maintained at a temperature that lies between the boiling points of the reactants; that is, at a temperature at which one of the reactants, such as ferrie chloride, is a gas and the other reactant, such as magnesium, is a liquid.

Figure 1 shows my process wherein product metal particles dispersed within a stream of molten reducing agent chloride, produced in a separate process, are caused to ilow into my separate, but contiguous, temperature build up and depositing zones.

Figure 2 shows my process wherein said stream of molten reducing agent chloride containing product metal particles is formed in a reaction zone that is separate, but contiguous, to said temperature build up zone.

Figure 3 shows my process wherein said reaction zone is divided into two separate, but contiguous, reaction zones.

Various methods are available for founing the stream of molten reducing agent chloride containing product metal particles employed in the process of Figure l. My U. S. Patent No. 2,647,826 discloses one such method. Another such method was disclosed by Maddex and Eastwood in the Journal of Metals, vol. 188, No. 4, April 1950. Other methods will occur to those skilled in the art.

Figure 1 shows my temperature build up 33 and deposltlng zones 35 cooperating to separate and accumulate 1n consolidated form the product metal particles (iron) entrained in a stream of molten reducing agent chloride (magnesium chloride), said iron bearing stream of mag nesium chloride being a product of a separate reaction zone (not shown) in which iron chloride is reacted with magnesium. The products of a reaction between one of the subject metal chlorides and one of the subject reducingagents will hereinafter be referred to as: the'v reaction In Figure 1, a stream of molten magnesium the top of an elongated chamber to form a free falling stream of reaction products 31 within said chamber, said free falling stream being either continuous or` a stream of discrete drops of molten magnesium chloride containing iron particles. As said stream of reaction products falls thru said chamber it ows into my temperature build up zone wherein the falling stream is mixed with a concurrently flowing stream of a heating gas 30 formed in a refractory reaction vessel 29 between magnesium 27 and chlorine 28:

Reaction 1 releases a much larger amount of heat than is necessary to maintain the product chloride in the gas state; the heat content of the product chloride of reaction 1 in excess of that required to maintain said product chloride in the gas state is heat available for the temperature build up phase of my'process.

In the temperature build up Zone of my process, the available heat content of said heating gas is employed to evaporate the molten magnesium chloride content of the falling stream of reaction products to form a heated stream 34 consisting of iron particles entrained in a stream of magnesium chloride gas, said heated stream being at a temperature that lies above the boiling point of magnesium chloride under the conditions obtained between the temperature build up and depositing zones of my process. The amount of heating gas required to achieve this end will depend upon the design of the vessel that forms said chamber-to be discussed later; the amount and temperature of the reaction products being flowed into the temperature build up zone; and the available heat content of the heating gas. In any case, suicient of said heating gas is mixed with the falling reaction products within the temperature build up zone to form therein said heated stream.

The heated stream leaving the temperature build up zone is immediately flowed into a contiguous depositing vzone 35 wherein said heated stream is flowed downward into contact with a surface 36 of a mass of the product metal, said surface being maintained at a temperature that lies above the boiling point of the magnesium chloride content of said stream by means of the heat content of the heated stream impinging into contact with said surface. As said heated stream is impinged into Contact with said surface, iron particles in said heated stream are thrown into contact with and deposit on said surface.

The heated stream, containing such iron particles as are not deposited on said surface, is flowed from said surface out of said depositing chamber by means of an outlet positioned at a location that is suiciently remote from said surface to prevent the outflow of the byproduct gas stream from the deposition process from interfering substantially with the downward flow of the heated stream as said heated stream moves from the temperature build up zone to the depositing surface. In View of the large volume of gas that must be handled in my process, per unit weight of deposited metal obtained, the outlet (not shown) in the wall of the depositing chamber must be large enough or numerous enough to prevent an excessively high gas back pressure from building up within my process vessel, not only because of the complications that such high back pressure creates in the Idesign of the vessel, but also because of the inuence of high pressure on the boiling points of the components of the heated stream, etc.

The byproduct gas stream leaving the depositing chamber outlet may be recovered or disposed of in any convenient manner. The preferred method of handling the byproduct gas stream is to introduce an excess of chlorine into it after it leaves or just before it leaves the `depositing chamber and then separating the resulting mixture of ferrie chloride and magnesium chloride by fractional condensation procedures; the ferric chloride being re-cycled back into reaction 2, below, the magnesium chloride being re-cycled back into the electrolytic process wherein the magnesium and chlorine are produced. Other methods of handling/utilizing the components of the byproduct gas stream will occur to those skilled in the art.

The reaction product stream being introduced into the 'generaba temperatur "b'nild'fup yzoneor my processf-wapresulhably Hforme'dfln afreactionrwhere `z'say, :5% #-ofireither zferricz chlorides or magnesium.:l When fthe reaction. lshcarriedout witha'crstoichiometricuexcess fof. 'ferricfchloride the: reaction product stream'zwill contain'ferrous-chloride;-` when-I itr is carriedffoutvwith a i stolchiometric r excess of magnesium; .the reaction product stream: 'will-A contain magnesium; f Reaction' 1-may;be

. carried-out with a.stoichiome11ic excessf ofeithen'chlonner4 ori magnesium,-v 1such lexcess, appearing in.` ythe-heating stream. Theheated: stream` emerging fromemy`V tempera- .-"turel build up. zone will reflect.' equilibrium betweenifall reactants, including suchreactant'excesses. For example, 1f.- reaction I1: 1s; carried with wstoichiometricfexcess of Other eifects 'arising' from lthe 4`use'ofstoichiornetric excessesZ taf-reactants in reactions 1 and 2 will occur to thosev skilled inthe art.

. 4As'fthe iron particles are deposited-on' -thezsurface'of the collectlng mass within my depositing chamberplsaid surface will tend to rise towards-the-'temperature build f lup zone, an'action that is prevented, in my-p'referred em- `I DQ'diment, 'by withdrawing the mass forming'said depositlingsurface ldownward at theI samerate thaty said, surface s rssbemg built up by the. depositing ironI particles, tthus :maintainingathe distance between thevtemperature Ybuild nupzone and-the depositing surface at'afsubstantially-con- 1 stant. value.

. :"lhe depositing: surface of my process-mayr bev eitherv 35 -solid iron ormolten iron, the latter being my preferred embodirnenh as shown in Figure 1. Whenthe depositlng surface isthe'surface 36 of va pool 8 of'molten iron, .fsaidfpool is retained: in a water-jacketed mold12"4 of` the continuous type Within which the molten depositing suri' 40 face iszmaintainedi and supported byI iron 37. that-,has been solidified atiuterface9 bythe cooling'action of said mold,

. the positionof'saild surface withx'respect to the Itempera- .pture buildup zone being 4controlled by withdrawing the -zsupportmg mass' of solidified iroiroutn of the bottomv of -lsaid'moldatthe'same rate Athat said surface tends to'` rise 1n saldmold asthe result ofthe .depositing process ytaking placefon said. surface., .'Thelwithdrawal' of the solidiiied .-2 iron out" ofthe bottom of said mold lmay be' controlled-vin -anyiconvenientmannen such as by one or more-sets of pinch rolls, :such as is conventional inmolds of this'type.

FThe vprocess of Figure l may be employed to'separate Y and'consolidate ,the product metal particles produced in `:reactions vinvolving a; chloride of a metal selected from .fthe 'group'.consisting. of, in addition to iron, chromium, lvanadium, nickel, cobalt,;.titanium and zirconium that' is .treducedto 'yield the subject productmetal by meansof a reducing agent selected from the group consisting'of i magnesium, sodium, potassium', lithium and calcium.

i Andsaid process may-be employed to produce fa' consolidated'metal product consisting ofan alloy inwhich one of the listed product metals is the base or predominating metal. The product metal producing reaction involved m forming the reaction product-stream of Figure 1 is not i a. part of .the invention constituting'the disclosure of i `,Figure 1,'however, in' order to be employable inztheprocness ofj Figure: 1,-'.saidxproduct metal producing reaction mustyieldf or'bymeans of the addition of a meltingI step must: bev-.capable of "yielding 4the reactionproduct stream oft'ligure 1,...said: reaction product streamconsisting of themolten 'reducing agent; chloride from. said .-reaction,

. `together with the .product .metalparticles dispersed therein, it being understood that 'the reaction product stream .from such .metal producing reactionsmay also contain 'the reducing-agent or a chlorideo'f the product metal,

depending upon thef'mannerin which the metal'producingreactionis.A carried vout.

1: While' Figure-l shows'myprocess being carried outin a completelywater-jacketedlvessel'1f32 containing circulating ywater 7-, in order" reassure thecomplete absence ofcoutaminating refractories, air and hydrogen from theen- '-vironment of my -processg-good--1results may" bel obtained Without Agoing to--this-extremein-certain-caseswherein a refractory suitably resistant to theproductmetal is avail- 10* able. -Iri theproduction:of"-iron, 4for example,the tem- `per-ature build-up zone may-be formed -of a-ves`sel l lined with magnesia.2 It-is important, however, infsuch ay case,

- that-reaction1= becarriedoutv withv an excess of magnesium,

rvfor chlorine-is reactive Withmagnesia; The-ability' of- -thefprocess yof reaction l 1l toI achieve substantially complete product metal purity with respect to the aforementioned'nonimetallic impurities depends, in largemeasure, oil-'the purity'of the reactantsemploye'cl in yreaction 1-andthe purityA of the reactionproductl stream withrespect tothese -nonmetallics. '-Thus', the-ferrie chloride andmagnesium being reacted toyield thereaction *product'streammustfbe-free from these non-metallics, as

l' must-thechlorineand magnesiumemployed in reaction l.

2 In-this connection, all commercially-available magnesium seems to contain ymagnesium oxide,- magnesium ynitride and freed `'of these' impurities 'by-first subjecting it toavacuum e treatment at a-temperatur'ein the -vicinity of its melting 'point' to" eliminatehydrogen, andI thenV distilling the -magnesium away from the oxide and nitride in--avacuum at as-lowa-temperature as possible. p It is-a noteworthy feature ofgrnyprocess -that -the high-temperaturefheat yrequired 'for this idistillation is1 available as aby-productof--my processythe'heatcontentof the byproductgasstream 6 emerging-v from the product metalf depositing zonebeing availablel for this-purpose.' The distilledreducingagent, whether it bemagnesium,sodium,A potassium, lithium'or vchlorine reacting with themagnesium content'of-the reaction product stream when said stream is mixed with said heating stream.v Another'm-ethodinvolves--theintroduction of chlorine insteadv ofthe heating-streameinto the temperature build` up zone, so that the reaction" between said introduced chlorinel and the magnesiumcon` tent of the reaction product-stream within the temperature build up zone will gain 'the desired gain iii-temperature. It istherefore apparent that; reaction 1 may-be ,caused to yield the Adesired temperature b uild up* by a reaction that takes place'within or` outside of vsaid zone,

or by means of a combination of these reaction sites.

The product metal producing reaction by which the reaction product stream of Figure 1 is obtained may be carried out in any manner that will lyield a productmetal `particle size capable' of rapidly ,settlingout of a moving stream of gas, and this` particle size requirement, in vvturn, is predicated, in part, on the specific gravity' ofthe product metal being produced. In;other Words, a larger particle size is required with titanium than with iron.

I prefer to carry out the metal producing reaction within the same vessel as that in which l( am carrying out thetemperature buildup and depositing phasesjof my process. Figure 2 shows such, an embodiment. 1The process ofFigure l2V has the merit of being a synchronous process inwhich the metal' producing zone, the temperature build up zone and the depositing zone are arranged in a straight-line ow wherein each step, while being substantially separate from the adjoining step(s), is contiguous thereto, so that the production rate of each step conforms to the production rate of the step (s) adjoining it.

As shown in Figure 2, .this preferred embodiment involves positioning a water-jacketed conduit 44 containing cooling water i7 that is suitable for conducting a owing stream of fluids in a vertical position directly over a horizontally-positioned product metal surface 43, said conduit being arranged and adapted to conduct said stream towards, and so into contact with, said surface, and then employing said conduit as `a reaction chamber within which said three zones are arranged in chronological order.

In the uppermost of thesevthree zones, the reaction product stream 31 is formed by reacting therein the reactants of reaction 2; that is, magnesium 27 and ferrie chloride gas 40. This, the product metal producing zone, will hereinafter be called the primary reaction zone 49. Within the primary reaction zone, ferrie chloride gas and molten magnesium are reacted to form the reaction product stream that immediately flows into the temperature build up zone 33, said stream consisting of one or more substantially continuous streams of molten-magnesium chloride containing the iron in dispersed phase, or said stream may consist of a stream of discrete particles of molten magnesium chloride containing said dispersed iron phase.

In the temperature build up zone, the reaction product stream of Figure 2 is shown being heated to the desired temperature by means of a heating stream(s) 3i) obtained according to reaction 1 in a separate zone (not shown) said streams 30 being introduced into the vessel via refractory conduits 45. If the reaction product stream emerging from the primary reaction zone of Figure 2 is caused to contain magnesium by employing therein a stoichiometric excess of magnesium, said excess may be oxidized in the temperature build up zone, if desired, by causing the heating stream to contain chlorine-for example, by employing a stoichiometric excess of chlorine in reaction 1. The basic function of the temperature build up zone is to lift the reaction product stream from a temperature that lies below the boiling point of magnesium chloride, preferably from a temperature that lies below the boiling point of magnesium, to a temperature that lies above the boiling point of magnesium chloride, preferably above the melting point of iron.

The heated stream flowing out of the temperature build up zone 33 is immediately flowed into the depositing zone wherein iron particles within said heated stream are caused to separate from the gaseous components of the heated stream by impinging said heated stream into contact with the surface 43 of a mass of the product metal (iron), the temperature of said surface being maintained at a level that lies above the boiling point of magnesium chloride, preferably above the melting point of iron, by means of the heat content of said heated stream. The heated stream, having given yup particles of iron to said surface, now flows out of the depositing chamber as the byproduct gas .T16 by means of outlet(s) positions in and penetrating thru the wall of the chamber that contains said product metal depositing zone, the byproduct gas outflow being arranged so that the gas outllow does not interfere substantially with the flow of the heated stream towards the depositing surface.

The depositing surface is shown, in Figure 2, contained in and supported by a water-jacketed mold of conventional continuous type, said mold -acting to cause the depositing surface to seal and close the bottom of the depositing chamber. As in the case of the process of Figure l, as the iron particles deposit on the depositing surface, the solidified iron shape 41 is withdrawn downwards in order to mainatin the depositing surface that it supports at a substantially constant level within the water-cooled mold, said shape 41 being formed at interface 15 from pool 42.

Reaction 2 must be carried out within the primary reaction zone in manners which cause said zone to be maintained at a temperature that lies below the boiling point of magnesium chloride (169l K.), preferably at a temperature that lies between above the melting point of magnesium chloride (987 K.) and below the boiling point of magnesium (1399 K.). The lower the operating temperature of the primary reaction zone is maintained within this range, the larger will be the size of the product metal particles produced therein, and the more efleient will be the product metal collection process taking place within the depositing zone.

The operating temperature of the primary reaction zone may be controlled by controlling the amount of reactants being reacted therein within an interval of time; by controlling the cooling elfect of the wall of the vessel containing said zone; by feeding an agent that is inert towards the reactants and resultants, vsuch as helium or argon, into said zone; by controlling the temperature of the reactants being fed into said zone; by controlling the size of a stoichiometric excess of magnesium being fed into said zone; or by controlling a combination of these controlling expedients.

When the primary reaction zone is being operated with the view of maintaining the maximum production rate` lthe usual case, or when one of the more active reducing agents, such as lithium, is being employed, ordinary methods of holding down the operating temperature of the primary reaction zone will not suffice. Under such circumstances, I hold the operating temperature of the primary reaction zone within the desired limits by increasing the size of the stoichiometric excess of magnesium being fed into sai-d zone. For example, if the operating temperature of the primary reaction zone is being held at, say, 1380 K., and for one reason or another the temperature of said zone begins to mount towards the boiling point of magnesium (l399 K.), I act to offset the rising temperature by increasing the amount of molten magnesium being fed into said zone until said temperature has fallen back to said l380 K. The extra, heat quenching magnesium added in this manner may be oxidized in the temperature build up zone by feeding chlorine into said zone, or said extra magnesium may be recovered from the byproduct gas after it leaves the depositing zone. The use of molten magnesium in this manner is particularly effective when it is being employed to hold the primary reaction zone operating temperature below the boiling point of magnesium, due to the temperature rise barrier: the heat of vaporization of magnesium. When, on the other hand, the temperature of the primary reaction zone begins to fall below the desired operating temperature, i act to lift the temperature to the desired level by cutting back on the amount of stoichiometric excess of magnesium being fed thereinto.

The metal producing zones of processes such as this are not restricted to the reduction of ferric chloride, of course, there being certain advantages in employing therein the reaction wherein:

The preferred primary reaction zone operating temperature for reaction 3 is between above the boiling point of ferrous chloride (1299 l.) and below the boiling point point of magnesium (1399 KJ, however, any temperature up to below the boiling point of magnesium chloride may be employed. The ferrous chloride of reaction 3 may be produced in any convenient manner, such as:

2FeCl3-l-Fe-9 3FeCl2 (4) @meseta Reaction: Sg of cour'seeyieldsialmixtue otchloridesgnhowf lever, lstlcharemixturedmay zbe reduced-` aoco'rdingitoire- -Faction-:Sfte= yieldvthef-reaction product stream: of my procsubstantially'll vtree"4 off 'irony a f substantial'4 excess f of ferrie chl'o'ridc over?4A the' indicated; stoichiometric requirements mustebe"fed'finttrthe-reaction area 4'whereI reaction 5' is "being conducted. whenfiii'thevpresence of a=-s`to`ichio -"addition` of fthe'extra' redctiontstep does not interfere with the highlyedesirble" synchronous 'procedure as' outlined- -if=Figure 2'.

In'Figure 3, magnesium Z7 is reactediwith yavt-stoichionesiurn is fed into the primaryireactiontvzone.tez-'provide :lathe: temperature -zbuild'r upz=zonerwitht .the:.magne`sium required: byfreaction: .1',-: .reaction v 1*.' being; :carriedcnuttwithin the temperature build v up zone: itself fby'ffmixing'; chlorine,

'Te 'bypfodilctchiorde -As wasfmentioneld-i previously, fthe; reducing; .-agentrch'loride 1 oth-the: reaction-1 product; ,stream} land? @the-reducing agentwchlorider ofc-reaction l: rleaveethe :depositing zone' infthetfor-m of gasgrthatg. isfgas.- 6, =f,16 an'd124;` respectively. .--lDepending -uponwlrow reaction. l isrcarried'iout, this-byw productchloricle -.gas .will contain, infadditionr to; therein-lo- Jride. otftheereducing Vregent, uncollected ,pjarticlesfA offthe :1 produc-twme'tal: andrunreactedrreducing agentin gasphase', 'onsaidfbyproduct will. containv a chloridey ofirthe product mct-al-.-.in the tlatter-case-wsaidr gas 1 may-also fcontaintuncolflected productffmetalwparticles. zIhreei thingsL are, of. iny .daresti-concerning Jthisz-:bYP-roduct' gas: f (1)1 its-.fheaLcon- 1 tent, (t2) -the-reducingi jt agent: f chloride: that it. conta-ins,

and (3) its .metal contentmEiiciency wouldrsugg'est -that the heat-content ofthe;I byproduct chloride-ga'sfbefutilized in the -Y overall: `'process ffor. some usefull purpose,-nsuch.;` as ythe :distillation .of-the; reducing gagent; in' orden to ;purify rine are employed in large quantities in my process, it t `would seem apparent that the reducing agent chloride byproduct must be treated electrolytically to recoven-'the 1 :reducing agent and chlorine for recycling baclcintore- .l actions 1 and 2. lIn order that this may be carried.out,;;th' -fbyproduct chloride gas must be freed from itsmetal content In my preferred embodiment, the heat of.vaporiza mtion'of the reducing agent chloride gas is.utilizedffor l -di's'tilli'ng the reducing agent, the distill'ingI proces'srbeingy caused to yield, in addition to the puriii'edN reducingagentf' .molten reducing agent chloride containing...limpurities, such as particles'of the product metal and/or a chloride 1' of' the -productme'tal. If lthe condensed reducing 'agent "chloride isatra :temperature thatV lies abovefthefboiling point'V of the reducing-agent, any= 'reducingagent present the byproduct' chloride-gas 'maybe separated-from the molten reducing agent chloride by-merely drawing the-reducing' agent' gas ''oifi ff'VVith the reducing agent chloride f f' lcollectedl inthe formof a molten pool, I 'purify it=and recover the productimeta'lfrom itl-bypassing 'chlorine "75 metric excess of ferrie chloride 40 in the uppermostfz:

.,zziZS; withrtlrereaction produce stream? asfit'. enters :said'tzonel' j 35i 10 into and thru' the moltenrpool, said chlorine acting t oxidizex'-therproductsmetalfparticles-:and such lowen prod- Lucttmetal: chloride,(s:), s-as.` may n bel Apresent to. the. higher, volatile. productefmetaltchloride, z suchas, :ferrie .chloride, 51ititanium:tetrachloridea'zirconium tetrachloride, etc. .The

vaporous productl'metalf..chlorides produced by :this chlos rine; oxidationfare .withdrawn from the.V molten :pool tothe i;ascondcnsedf:separateronI-r said: molten'` pool,; Aand so: reen covered dffonfrecynling.el-Jack: intot ithe product; metal producing reactionitthezzpnried areducingsiafgent chloride-2 of msaidnnolten pool being passed to the electrolytic process efwherein. thev reducingagent and chlorine are produced. t'. Theproduct metal content of the byproductrohloride :fugas emergingt from the depositing zone may-.alsowbe recovered from said gasby. merely feedingnough-:chlo- 1 nonine into thel byproduct chloride gas after said gasemgerges 1: iirom the depositing zone to oxidize theiproduct r-tmetal particles to the highest state of oxidation -obtain- .ab1e.at.the temperaturestinuolved; .that is in. the case of 2o `iron,ferridchllariile,,.or.- in thecasel of titanium, ,titanium .tetrachloride.. Withlthelproduct:metal oxidized, .the product Ivrnet`alr1iilil-rideproducedv thereby is. separatedfrom `,the. reducing,T agent .chloridelbyr fractional condensation.

u Alloys 5 "'lfWiiife'EI-have/Iempl' yedfthe reaction wherein #an-iron --vificli'l'orid" i 'due'eddwitht magnesium to' illustrate .how H my-'processi iurict-iensg as'f m'entioedl ypreviously; 'other f-metalsmaybe-poduedby my' proces-stand other red'ucq ing-#agents-may' beremployed Inrgene'rahf' myf-process is applicable iti\those":situations v wherein ametal selected 'fromtltegroup'corrsi-sting of- .-iron', chromium; vanadium, ffnickelgecob'altgttitanium and zirconium i'sformed bye-reracting-fatll'chl'oiide osaid/met'a-l'with a'reducing ag'ent' that :is isubstantia'lly morelelectropositvel ltharrvsaidiametal. JSuitable=ieduuing1agents for` tli'e'flis'tednproduct Jlnet'als'are:

.2 magnesium@sodiumsfplotassium, lithium;r calcium; Y' et'c:

Withl x`few'exceptions' the listedlfproduct'metalsv farei'not utilized L to' 'anyfgrea't-f entente r'rttheiri ipure- `ormf,--z 1being .rf-usually employedritifffhe'uform ofall'cys infwhichasaid 40:? metali is the predhminatingrmetal present-r "Myf-'process isfparticularlyfisuited to-the-v production f of these'ialloys. "-'"While tlrerefareAY '-aLnumberr of'r approaches vtoifthe produc- "tion" of.z an@ 'alloyt in1-myHp'r-ocess; it:-s'lrou`ltl` :always-A betthe ,1g-faim to' causeeachsmetalfparticle ini-the.'reaction-product 4"stream to conform in composition to ltheldesiredl1 analysis. There being sotrnanyrl rtypes o'f fa'l-loys'fl of' t' the listed base metal'st. it wouldl-helimpossib'le toudiscuss.` the production of-eacl1 alloy specifically, however;l the basic approaches ,504 tofall'oy productioir-williehable thexuser to producer any of the more` e'ommorrfalloys'.presently-y beingwemployed inrindustry.

If:theffalloyingrelement isionefh'aving achloriderthat `is -reduciblerfbysrtheureducingfr agent=being lemploy/edit. to

it In viewofthe fact that-the reducing agent and chlo l'rprod'uce' the'base metalin'the primary' reactionzon'e; :then

:the desired? allolymay! be fproducedf byf reacting v"said-re- '1 4ducing` agent. With rasmixture of' chlorides. in sai-d f'pri'mary reaotionv-zonellftheicompositionf of '-the chloride#v mixture ..rflbeinglsuchfthatfits reduction-.yields particlesy of--the desired I, solfialloys? If.thefalloying-element isfsoluble inthe-'reducing agent, such `asi islftlrercasewhenr yaluminum is the?. alloying t element: and/.magnesium is the` reducingl agent; their i the desiredualloy is*advantageously.y Yformed by '-feeding' the x aluminumde'quiretdfzfortme alloy\-vformation fintox-tirez"primarylzreactionftzone alloyedwith"the-magnesiumthat is s u may: :be.fedrdirectlyrintofthe' poolfof -molten f'metal fthat 'fHform'st-the depositing surface'of'my'fprocess, however, it should be pointed out that this procedurewill-.notyield asf-uniform :recomposition asf-the'foregoingaprocedures.

Thc icrmat-icn of yiron-carbontalloysoiiersan interest- -ing problem infmy processf'however', as will be-fdiscussed flater; the-.redu'cti'onlfof a mixture`of-ferrio'chloride- Iand vcarbon,'tetraclrltiride,rlioth int-gas-rphase, will yi'eldf-'such .alloysm Ilm order tosfmoref'clearly -`disclose the operation gf.,- my: promener-thaiifollowingvspecic' 'examples l are" "given,

Examples In order to determine the efficiency of my process in forming an iron that wasl essentially free from carbon, hydrogen, sulphur and phosphorus, I chlorinated ordinary mild steel at an elevated temperature employing the general technique given by Lundell, Hoffman and Bright, Chemical Analysis of Iron and Steel, John Wiley & Sons, Inc., 1931, pages 413 thru 420, cylinder chlorine being freed from water for the purpose by freezing it out. The

mild steel had the following composition:

Percent Carbon 0.14 Manganese 0.42 Silicon 0.09 Phosphorus 0.019 Sulphur 0.027 Copper 0.18 Nitrogen 0.005

The test unit was a vertically-positioned pipe that was cooled by circulating water, much in the manner shown in Figure 2. The inside diameter of the pipe reactor was 2 and 3Ai, the length being 24" from the top to the depositing surface. The length of the primary reaction zone was fixed at by positioning the chlorine entrance ports 10 from the top of the reactor pipe, the chlorine entrance ports-2 in number-were formed of graphite pipe encased in stainless steel tubing, the steel tubing being water cooled in the area where they passed thru the waterjacket that formed the reactor wall. The. unit was prepared for operation by removing the collecting (depositing) surface from the depositing chamber and preheating said surface to a red heat, the starting depositing surface being composed of rammed magnesia in the tests involving iron-base metals. The magnesia, consisting of prefused magnesia grains bonded with magnesium chloride, was rammed in the form of a plug that would slip easily up into the open bottom of the depositing chamber, much in the manner that shape 22 is positioned in Figure 3, however, during the tests, the depositing surface was positioned lower in the water-jacket mold 23 'than shown in Figure 3, due to the fact that the tests were run on a batch basis; that is, the depositing surface was not lowered during the test runs in order to maintain the depositing surface at a fixed position within the mold, the deposited metal being allowed to build up until the top of the mold was reached or until the run was finished.

In the test runs involving the mild steel chlorides, I fed the mixed chlorides, in vapor phase, into the top of the reactor at the rate of about 125 grams per minute, at the same time that molten magnesium, at about l050 K., was fed thereinto at a rate of about 60 grams per minute. This ratio of product metal chlorides to magnesium was more than a two-fold amount of magnesium over the stoichiometric proportion required to merely reduce the product metal chlorides to yield the product metals. After the primary reaction had been in operation for a short time, during which the water-jacket wall of the pipe reactor became coated with condensed magnesium chloride, I began the temperature build up phase of the process by feeding chlorine into the temperature build up zone at the rate of about 80 grams of chlorine per minute. This amount of chlorine was enough to oxidize a substantial portion, but not all, of the magnesium content of the reaction product stream. As soon as the chlorine introduction was commenced, the magnesi depositingsnrface was inserted in the depositing chamber in thepath of the heated stream emerging from lthe temperature build up zone.

The byproduct chloride gas emerging from the depositing zone was exhausted from the depositing chamber thru an 8 length of 4" pipe. After a few minutes of operation, it became necessary to cool the exhaust pipe by v means of a water spray, care being exerted so that the exhausting pipe was not cooled below the melting point of magnesium chloride, so as to avoid clogging the pipe.

This exhausting arrangement condensed a portion of tht byproduct chloride gas, the balance being exhausted int( the air where it formed a heavy smoke. The exhaust ar rangement was effective in preventing air from diffusing into the reaction pipe, so that, once a few minutes opera tion had cleaned all hydrogen, oxygen and nitrogen ou of the reactor, the entire reduction and collection proces was carried out in the absence of these impurities.

After the depositing surface had been placed in posi tion, the reduction/collection process was continued fo.` 10 minutes; that is, as long as the supply of mixed chlo rides held out. During this 10 minute interval, the inpu of iron was estimated at about 430 grams. The weigh of the melted and solidified deposit collected on the de positing surface was 352 grams, reflecting a process yielr of 81.8%. In'spite of the absence of deoxidizers in tht deposited iron, the small ingot took a shrink, indicatinI the absence of oxygen therein. Analysis of the top o.r the ingot showed:

The low nitrogen level at the top of the deposit dem onstrated the ability of the process to clean up the re actors atmosphere, for analysis of the first iron deposited showed 0.006% nitrogen.

A thermocouple positioned in the primary reaction zone showed a fairly uniform reading just over 1400 K throughout the 10 minute run.

While the manganese chloride and copper chloridr. content of the mixed chlorides were left behind when thtiron chloride was distilled into the primary reaction zone, in other tests involving the reduction of the molten mixen chlorides with molten sodium, essentially all of the copper content and about 50% of the manganese content of the original mild steel were recovered in the deposited iron. In these tests employing sodium as the reducing agent, it was observed that sodium is not as efficient as magnesium in gaining the desired final temperature build up-this, in spite of the known higher activity of sodium. Thus, while 6 moles of magnesium would gain the de- 'sired temperature build up, 6 moles of sodium would not-3 moles being the equation requirement in both cases. With sodium, I was forced to employ 3 times the stoichiometrc reaction proportion of the reducing agent in order to gain a deposit -of the required type.

In another test run involving the use of my process foi the production of a very low carbon stainless steel of thc l8-8 type, I chlorinated an 18-8 to form a mixed chloride condensate. The composition of the 18-8 that was chlorinated:

The melted mixed chloride condensate and molten magnesium were sprayed into the primary reaction chamber in substantially the same proportions as in the mild steel test run. The test was operated for 21 minutes during which the primary reaction zone functioned at about 1400 K. The 18-8 deposit recovered from this test reliected 68.9% of the calculated metal input, Thel top of the 18-8 deposit analyzed;

V'Thefziilurejolf the silicon-andzirconium to be Aefliciently recovered was due to their. condition in the original 18-8,

for another-test 'demonstrated' that these two elements l could be eiciently recovered.' In this `latter test. I chlorinated a silicon-zirconium alloyof the following composition:

A Percent Silicon '50.11

Izirconium 37.78

` Iron '-11.10

InI this test the'proportions employed wereabout the '-same as in the previous tests, the reaction beingrinstigated `was l70.5

l yInlanother 'test' involving the production-of titaniumlin -myy process, I vchlorinated a1 ferro-titaniumof the-followging composition:

Percent Carbon 0.05 Titanium 41.55 Aluminum 6.70

f Silicon 4`.11 `Iron i 'Balance By fractionally condensing thechloride gas, Ifseparated most of the iron, the collected condensate containing mainly titanium and silicon tetrachlorides. The titanium and silicon chlorides-were sprayed in thel liquid form into a sprayed molten magnesium within the primary reaction 'zone in the following proportions: mixed-chlorides, 100 grams per minute; molten magnesium, 50 grams per minute; chlorine, 65 grams per minute.4- These rateslof feed were based upon the net feed duringthe run, 19

minutes, asfwere the feed rates in 'all oftmy tests,,however, a certain measure of rate of feedvariation occurred during the coursefof: the =run, due to variables which have no interest here. The yield of deposited metal during'" this I9minute-run was 72.8%, the composition of "1"; the melted deposit:

Percent Carbon Trace.

Titanium v88.9 (by difference). Silicon 8.50. Aluminum Trace.

Feeling thattheV introduction Y:of chlorine into contact with a metalproduct stream containing as active a metal "as titanium would oxidize a substantial portion of the metal particle content of the reaction product stream, I

arranged another test wherein I formed the vtest unit` in the shape of a Y, theI primary reaction zone occupying one arm, reaction '1 the other arm, so that the reaction product stream, formed with a 'moderate excess of magfinesium, would merge withl the heating stream at'the point where' the two.'` arms lmet to formnthebaseiof .theY.

.disagrees The process yield fromfthist'est was 78.5%, indicating i when chloriner was introduced! idirectlyinto theA temperature buildiup zone;

"Whilefthe E operating temperature (of, -the` primaryzreactionfzone wasi'nrtheyicinityof the boiling point' of the reducingagent ini'mostf'omy:ltests, in a` few testsithe --temperatureroseffwell oversaid boiling-.point,.and it is Y Inoteworthy that-v thei :higher` thentemperature yrose over this boiling point,=the-poorer. \the:vresults,.A in terms` of process yield.vv Onfthe other hand,` themere fact that .-therprimary reaction fzone `was,V operating at atemperature #over the-boilingpointof-the reducing .agentV should not be taken to mean that the reaction being entertained -therein-was alvapor/phase reactiom-ffor itwas clear throughout my work that substantially all ofthe product metalryield arose from a reactiombetweenthe product mmetaly chloride in; gas phasehand the`- reducing agent in yliquid phase.` -Tho-reasons why -largef productmetal ,parf =ticles couldl belproducedtina primary reaction zone operating at a1temperatureiwell.above'the.iboiling point of the reducing agent' wasy dueto thelag that occurred at the boiling'pointrof thereducingfagent, dueto thelatent l-fheat of vaporization; In other rwords,rasra' given'particle -of molten reducingagent.reacted.with a gaseous environment containingthe product metal chloride, the size of the.` mass of productmetal being produced continued to increase until v`thefentire system-reducing agent/metal f product/ reducing; agentchloride-rose over the boiling i point* of the -reducingf agent, lwhereupon further reaction between the reactants-in vapor phase yielded product metal -`particles' too-,ne to =be collectedon the depositing surface. The higher the operating temperatureVthe -quicker-thesystemI is lifted over4 this critical level, and Jthe lower the processyieldgihowever,` it shouldbe mentioned that acceptable' yieldsv'canbe`obtained so long as the operating temperature is maintained below the boiling Y point of the reducing, agent, chloride.

. of thetusual magnesia.

given.

During the course' of my- Work, it` occurred tome that my preoccupation with a low operatingtemperature inl the primary reaction-"zonef-might be unnecessary. Accordingly, Iran a strictly vapor phase testin a unit built somewhat in the vmannerof Figure 1. In ythis test, I fed molten magnesiumA andrlchlorine -into vthe`upper graphite chamber toform a highly heated stream consistingiof magnesiumchlorideegasand.magnesium4 gas', and then I owed this" highlyI heated streamrinto a primary reaction zone wherein ferric chloride gaswas'reactedy withA the magnesium gas content of the highly heated "stream, ythe product stream from-this*reactio`n thenvbeing owed -into contact -with.the depositing surface. Y The'yieldtrom this :test was 19.9%.

In` another test of thisvapor` phase-reaction, I reacted titanium' tetrachloridey with'.the y'highly heated stream containing magnesiumV and obtained apro'cess yield of 16.27 This test was of further interest due to the fact that-the starting depositing surface was graphite instead TheY bottom ofthe deposit ran 0.52% carbon, the`topran 0.17% carbon, and itfwas evident that this titaniumv deposit was ductile.

In View of the fact that the chlorination lprocedureemployed to produce my product metal chlorides always eliminated substantially all of -the carbon from the original metal, I experimented with the use of carbontetra- 'chlorid'e'for replacing this :lost carbon, due to the imporrtance' of the carbon/iron, etc. alloys.' I found that the introduction of c'arbontetrachloride into the primary reaction zone resulted in the formation ofthe desired alloys containing` carbon.

The reactor While my invention does not involve reactor design, certain details of such design will be mentioned, so as to permit advantage to be taken of my experiences. As has been mentioned before, the primary reaction zone may be formed of a non-metallic refractory if the product metal will tolerate contact therewith. The use of a rammed lining of magnesia in a primary reaction zone wherein an iron alloy is being produced being an example of this. Other examples will occur to those skilled in the art. I prefer, however, to employ a water-jacket vessel to contain the primary reaction zone, the temperature build up zone and the depositing zone, such as is shown in Figures 2 and 3.

In Figure 1, the product metal depositing zone isVV shown entertaining a depositing surface consisting of a molten pool of product metal 8, the molten metal of pool 8 being caused to solidifyat interface 9 by the cooling action of mold 12. As the product metal content of the heated stream collects on the depositing surface, the product metal shape that supports said surface is withdrawn in the indicated direction by conventional means (not shown) such as rolls. The product metal shape within mold 12 may be maintained in a controlled atmosphere by the introduction of fluid 10 in the indicated manner; that is, into the gap positioned between the solidified product metal shaped and mold 12, packing means 11 being positioned at the bottom of mold 12 in order to facilitate such atmosphere control. The reaction product streams are shown entering the heating stream. In passing thru the water-jacket and cooling water 7, said reaction product streams are naturally contained within suitable pipes (not shown).

Due to the cooling action of water 7 on the lining of the water-jacketed vessel, the reducing agent chloride deposits on said lining to form a heat-insulating layer that acts to minimize the heat losses in the process. There are a number of peculiarities concerning this deposit of reducing agent chloride. When a reactor rst starts up, the reducing agent chloride deposits on the waterjacket to form a more or less porous frost, however, as the operation proceeds, this frost layer becomes more and more dense until, finally, it consists of a fused layer that is solid next to the bare water-jacket and molten next to the heated stream. The molten phase of this deposit is in the process of being continually formed and then flowing down the Wall towards the molten depositing surface, and, unless the molten chloride is prevented from falling upon said molten surface, a continual series of explosions will occur in the reactor. While there are a number of methods for preventing this condensed, molten reducing agent chloride from falling on the depositing surface, I prefer to handle the situation as shown in Figures 1, 2 and 3. In Figures 1, 2 and 3, the depositing surface has the same shape and size as the cross section of the temperature build up zone in the same plane. In this connection, it is important that no dimension of said surface be larger than the corresponding dimension of the cross section of the temperature build up zone, and, while not shown in the figures, it is preferable that each dimension of said surface be smaller than the corresponding dimension of said zone. Under these design circumstances, the molten reducing agent chloride will flow into the byproduct chloride gas chamber 14, either by entrainment in that portion of the heated stream that Hows into chamber 14 without contacting the depositing surface, as shown by the flow lines in the figures, or by falling upon an inclined top 13 of the mold 12 containing the depositing surface, or both. The molten reducing agent chloride that flows into chamber 14 will quickly evaporate off, or it may be withdrawn in the molten state by means of an outlet (not shown). In the event that this molten chloride is removed frorn chamber 14 by evaporation, the product metal particles which it contains will remain behind in chamber 14 to form accretions. Such accretions, and such other accretions of the product metal as may occur at undesired locations within the reactor, may be periodically removed by feeding chlorine or a reducible gas, such as ferric chloride, titanium tetrachloride, etc. in large excess into the atmosphere of the reactor for a short while. If desired, the product metal accretions within chamber 14 may be removed by feeding chlorine, ferrie chloride, titanium tetrachloride, etc. into said chamber 14 by means of a suitable inlet (not shown) in said chamber 14. On the other hand, the development of such accretions within chamber 14 may be prevented by continuously feeding a substantial excess of such oxidizing gas into chamber 14, this procedure being compatible with the byproduct reducing agent chloride recovery and purification procedure previously disclosed.

The reactor of Figure 2 offers some design problems that differ from those of the reactor of Figure 1. In Figure 2, the heating stream is formed in a reaction chamber (not shown) that is separated from the reactor itself, the heating stream being flowed from the chamber in which it was formed into the temperature build up zone by means of suitable, refractory-lined pipes. The refractory lining of such pipes is preferably graphite, although other refractories will occur to those skilled in the art. Chlorine 28A is shown being introduced into the top of the reactor of Figure 2. This is an optional arrangement for gaining a preliminary temperature build up in the primary reaction zone or for periodically purging the reactor in order to remove accretions. Water-jacket 17 is shown forming the three contiguous phases of this synchronous process: the primary reaction zone, the temperature build up zone and the depositing Zone, the depositing surface being formed by pool 42, pool 141 being supported by the iron shape that is formed at interface 15. The byproduct chloride gas 16 of Figure 2 is withdrawn from the vessel in the manners previously described. Inert gas 18 is shown being fed into the gap between the iron shape and the mold that formed said shape, packing 19 being employed to retain gas 18 within said gap.

Figure 3 pictures the use of chlorine in the temperature build up zone and the use of a separate, but contiguous, dichloride zone wherein the ferric chloride is rst reduced to ferrous chloride, as previously described. The direct introduction of chlorine in the hot temperature build up zone requires certain precautions if attack on the vessel is to be avoided. In general, the chlorine must be led into said zone by means of cooled graphite pipes or other suitable refractory pipes. In Figure 3, depositing surface 2t) is shown being supported by shape 22 that is formed from pool 48 at interface 21 within mold 23. The entire vessel, including that portion containing byproduct chloride gas 24, is shown formed by a suitably designed water-jacket 51 containing cooling water 25.

M iscellaneous factors While I have described the various embodiments of my process wherein the heat required in order to maintain the depositing surface at a temperature that lies above the boiling point of the reducing agent chloride being formed in the primary reaction zone is developed without the use of supplementary heating expedients, it seems plain that such supplementary heating may be employed, if desired. For example, the conduction of heat away from the depositing surface by ow downward along the product metal shape may be offset by inductively heating that portion of the product metal shape which extends below the mold. Other heating expedients will occur to those skilled in the art.

The Kelvin temperature scale is employed herein because it is the scale employed in Brewers Tables: 3.7, 3.8, 6.3, 7.1 and 7.2, Chemistry and Metallurgy of Misc. Materials, McGraw-Hill Book Co., 1950, these tables being the best and most complete gathering together of data concerning the melting points, boiling points, etc. of the metals, reducing agents andchlorides herein involved. In this connection, when I refer to the melting and boiling points of the various metals and chlorides herein involved, I naturally means Ithe melting and boiling points exhibited by these metals, reducing agents and chlorides under the conditions which are characteristics of my process. One of these conditions is the relatively high pressure under which my process operates, due to the sudden and large expansion of the reducing agent chloride from a liquid to a gas within the confines of my process. Just how high this process pressure will be, in any given operation, will depend upon how the process is being operated and upon the back-pressure arising from the design of the outlet that withdraws the byproduct chloride gas from the reaction/collection vessel. This process pressure naturally has an elevating eiect on the melting points and particularly the boiling points of the process reactants and resultants. Whatever the process pressure may be, however, when I refer to the boiling point of magnesium chloride, for example, I am referring to the boiling point exhibited by magnesium chloride under the environment obtained within my process vessel.

While I naturally prefer to operate my process in the absence of air and hydrogen, for the reasons previously given, the presence of these gases within my process is permissible if such is desired.

By denition, the reducing agents of my appended claims are those metals which are suciently electropositive towards the subject product metal chloride to reduce said chloride with the formation of said metal; magnesium, calcium, sodium, potassium and lithium being examples of such reducing agent metals.

While I prefer to operate my heating process for the purposes hereinbefore stated, my heating process may be employed to merely heat treat the product metal patricles contained in the subject reaction product. In other words, my temperature build up zone may be employed Without the subsequent collection of the heated product metal particles in a consolidated form.

Having now described several forms of my invention, I wish it be understood that my invention is not to be limited to the specic arrangements hereinbefore described and shown, except insofar as such limitations are specied in my appended claims.

I claim as my invention:

l. In the process wherein a chloride of an element selected from the group consisting of iron, chromium, vanadium, nickel, cobalt, titanium and zirconium is reacted in a reaction zone with a reducing agent metal selected from the group consisting of magnesium, calcium, sodium, potassium and lithium to form a reaction product containing a product metal in which at least one of said elements is a major constituent and a chloride of said reducing agent metal, and in which said product metal is then separated from said reaction product, the method of separating said product metal from said reaction product, which comprisesr flowing said reaction product to form a reaction product stream in which particles of said product metal are dispersed within a molten stream of said reducing agent metal chloride, the temperature of said reaction product stream being in the range between above the melting point and below the boiling point of said reducing agent metal chloride; owing said reaction product stream into a temperature build-up zone that is essentially separate from said reaction zone; reacting suf- 'licient chlorine gas with sufficient of said reducing agent metal in fluid phase to form a heating stream of a gas consisting essentially of a chloride of said reducing agent metal at a temperature that lies substantially above the boiling point of said reducing agent metal chloride; mixing said reaction product stream with sufficient of said heating stream within said temperature build-up zone to lil lift the resulting mixed stream to a temperature thatv lies above the boilingpoint of said reducing agent'metal chlo' ride; owing said mixed stream into a contiguous depositing zone wherein particles 'of the product metal content of said mixed stream are brought into contact with and collected on the heated surface of a mass of said product metal by flowing said mixed stream into contact with said surface, said surface being maintained at a temperature that lies above the boiling point of said reducing agent metal chloride with the aid of the heat content of said mixed stream; and then flowing the gas content of saidl mixed stream away from said surface.

2. The method according to claim l in which said reaction product stream contains said reducing agent metal and in which the temperature of said reaction product stream is lifted by reacting the reducing agent metal content of said reaction product stream with chlorine gas within said temperature build-up zone.

3. The method according to claim 2 in which said reaction between the reducing agent metal content of said reaction product stream and chlorine gas lifts said reaction product stream above the boiling point of said reducing agent metal chloride.

4. The method according to claim l in which said heating stream is formed by reacting said reducing agent metal with a stoichiometric excess of chlorine gas.

5. The method according to claim l in which said heating stream is formed by reacting chlorine gas with a stoichiometric excess of said reducing agent metal.

6. The method according to claim l in which said reaction product stream is formed .in said reaction zone and is then owed downward into a contiguous temperature build-up zone wherein suicient of said heating stream is mixed with said reaction product stream to lift the resulting mixed stream to a temperature that lies above the boiling point of said reducing agent metal c1110- ride; in which said mixed stream is then flowed downward into a contiguous depositing zone wherein said mixed stream is flowed into contact with said heated surface to collect a substantial portion of the product metal content of said mixed stream on said heated surface; and in which the gas content of said mixed stream is then llowed away from said heated surface.

7. The method according to claim 6 in which said reaction zone is divided into two contiguous reaction zones, the rst of which being employed to conduct a reaction between said reducing agent metal and a stoichiometric excess of said chloride of said element, said chloride being oxidizing towards said element, to form a first reaction zone reaction product stream consisting of said reducing agent metal chloride, a lower chloride of said element and said excess of said oxidizing chloride; and in which the reaction product stream from said iirst reaction zone is then owed downward into a contiguous second reaction zone wherein said rst reaction zone reaction product stream is reacted with a stoichiometric excess of said reducing agent metal to form a second reaction zone reaction product stream containing particles of said product metal dispersed in a molten stream of said reducing agent metal chloride, said second reaction zone reaction product stream being then flowed downward into said contiguous temperature build-up zone.

8. The method according to claim l in which said heated surface is maintained at a temperature in the range between above the boiling point of said reducing agent metal chloride and below the melting point of said product metal.

9. The method according to claim l in which said heated surface is maintained at a temperature that lies above the melting point of said product metal.

10. The method according to claim l in which said product metal consists of an alloy of said element with at least one other element and in which said alloy is pro'- duced 1in said reaction zone '.by reacting a fmixture `of chlorides Withsaid reducing -age'nt metal.

'11. Thefme'thod according to claim 1 in which 'said heated nsurfaceas thefsurface of a pool of a.mo1ten alloy consisting-of saidelement and at least oneotherelement and iuwhich said other elementis introduced into said pool byfeeding ysaidy other. element;A into said pool.

12. The method according to claim 1 inwhichmsucent'chlorine gas is introduced .into the gals'stream frafter it has'owed away from said'heated 'surface to oxidize product metalsparticlesentrainedn `said byproduct chloride gas stream to the highest state of oxidation obtain- 2Q able at -the Lteinp'erature Iinvolved, amd `in which 'the "resltingsstream. is then ifractionally condensed fto separate the components :'thereof.

References Cited in the "leof his'patent UNITED STATES PATENTS 2,205,854 Kroll June 25, I1940 2,668,750 Krchma Feb 9, l'1954 FOREEGN .PATENTS 505,801 Belgium ?Sept, 29, 1951

Claims

1. IN THE PROCESS WHEREIN A CHLORIDE OF AN ELEMENT SELECTED FROM THE GROUP CONSISTING OF IRON, CHROMIUM, VANADIUM, NICKEL, COBALT, TITANIUM AND ZIRCONIUM IS REACTED IN A REACTION ZONE WITH A REDUCING AGENT METAL SELECTED FROM THE GROUP CONSISTING OF MAGNESIUM, CALCIUM, SODIUM, POTASSIUM AND LITHIUM TO FORM A REACTION PRODUCT CONTAINING A PRODUCT METAL IN WHICH AT LEAST ONE OF SAID ELEMENTS IS A MAJOR CONSTITUENT AND A CHLORIDE OF SAID REDUCING AGENT METAL, AND IN WHICH SAID PRODUCT METAL IS THEN SEPARATED FROM SAID REACTION PRODUCT, THE METHOD OF SEPARATING SAID PRODUCT METAL FROM SAID REACTION PRODUCT, WHICH COMPRISES: FLOWING SAID REACTION PRODUCT TO FORM A REACTION PRODUCT STREAM IN WHICH PARTICLES OF SAID REMETAL ARE DISPERSED WITHIN A MOLTEN STREAM OF SAID REDUCING AGENT METAL CHLORIDE, THE TEMPERATURE OF SAID REACTION PRODUCT STREAM BEING IN THE RANGE BETWEEN ABOVE THE MELTING POINT AND BELOW THE BOILING POINT OF SAID REDUCING AGENT METAL CHLORIDE; FLOWING SAID REACTION PRODUCT STREAM INTO A TEMPERATURE BUILD-UP ZONE THAT IS ESSENTIALLY SEPARATE FROM SAID REACTION ZONE; REACTING SUFFICIENT CHLORINE GAS WITH SUFFICIENT OF SAID REDUCING AGENT METAL IN FLUID PHASE TO FORM A HEATING STREAM OF A GAS