US3186833A - Method for reducing copper oxide - Google Patents

Description

June 1, 1965 R. E. cEcH METHOD FOR REDUCING COPPER OXIDE Filed Jan. 2, 1963 Sheets-Sheet 1 ---22 Copper Oxide 23 l I I Sulphur Containing Alta/l or AIM/in: Reducing Agent Ream Earth Chloride I 24 Separator By-Products l 25 Copperar Copper Oxide Scrubber 50 Poe/red V l Tower 1 6/ Oxidizer LLi Reflux Distillation 6'2 Ammflfll'fl Condenser Carbonole 47 s, 49 coppef orage .Scr p ""32 55 3 7 West:

Pressure. 32'- Filler 9= =l In venfor: Rob err E Cec His Afforney.

June 1, 1965 R. E. CECH METHOD FOR REDUCING COPPER OXIDE Filed Jan. 2, 1963 5 Sheets-Sheet 5 Ms mm m r I m U r In B W m/ H n m w F 4 .m r ll 0 0 0 0 m m w w m m 5 4 2 m y lmu a 93mm m w m m o m a s 4 2 5 m E= o :35 n m 0 m T I 5 m6 m w M n m I I M .7 e d me w W I H I m m e n 2 mm m n F .m .9 1 .H a a m 8 I .I 3 .m r 9 1 T I m I u m 0 o 0 0 a m a 4 2 0 w w w M 6 5 4 2 m l 0 a l 5 5 s m 0 a. 4 I M m a 0 r 5 3 3 m 5 U m] :0 5m 2 .w I l .0 4 3 m n u m m l W a u u a u a w w w 4 m a lw m SE R musm 2 W m 0 2 2 I 5 I 0 0 m a a w Time of Temperature-Minutes Time a! Temperature-Minutes In ve mor Robert E. Cec

His Afforney- June 1, 1 965 Filed Jan. 2, 1963.

5 Sheets-Sheet 4 Fig. 7T

0 m w n w 5 6 .9 H 0 2 I a a B w ll 0 0 u 0 a 0 0 a w 4 2 0 I H S= 3 39 5.

u n m x a m :2 m m -2 w. m h :0 3 4 l I 2 a E 6 ll5 .m 7 p 0 v 0 m w w w z a s n 0 M -mm .n e M w l a m r w n m n n ..m.m I C a a P e R m T a a 0 0 0 0 6 8 6 .7 M m M 6 6 Q Q PEQE s m 0 m WW I? w 9 e n 7 f B 3 M m E w 5 a 0... B m 0 0 0 0T M 8 w 4 2 0 0 a w a m w 6 4 2 I0 20 7 Time at Temperature-Minutes lnvemor: Robe/2 E. Ce /7,

His Afforney June 1, 1965 R. E. cEcH METHOD FOR REDUCING COPPER OXIDE Filed Jan. 2, 1963 5 Sheets-Sheet 5 Fig. 23.

Fig. 24.

Inventor Robert E. Cec y W a His Attorney.

United States Patent 3,186,833 METHOD FOR REDUCING COPPER OXIDE Robert E. Cech, Madison, Wis., assignor to General Electric Company, a corporation of New York Filed Jan. 2, 1963, Ser. No. 249,036 15 Claims. (Cl. 7572) This invention relates to the metallurgy of copper and more particularly to new and novel processes for obtaining purified copper oxide or metallic copper from either ore or scrap copper sources.

Copper remains one of the most important metal in recorded history, its use according to some authors dating back some 6000 to 8000 years. Such early use of this metal was undoubtedly due to its frequent occurrence in the native or metallic state, since the concentration and smelting of copper ores are essentially products of the past 200 years. Progress in the development of copper smelting furnaces was extremely slow initially but ultimately did gain momentum as the need for copper grew beyond the point where native copper and crude furnaces could supply the quantities required by expanding industries. As copper consumption multiplied, ores containing as little as one percent copper and less were mined, concentrated and the copper values extracted. Additionally, copper scrap became economically valuable. Unfortunately, copper obtained from ore and scrap sources in almost all instances contains other metals alloyed with it which must be removed from the copper either because the alloying metals are valuable or because they render the copper unsuitable for the intended use.

Among the metals found in copper ores are iron, nickel, silver, gold, platinum, palladium, osmium, iridium, ruthenium, rhodium, molybdenum, cobalt, lead, zinc and arsenic. Processes of varying degrees of usefulness have gradually evolved for removing these and other impuri ties from copper. For example, sulfur, zinc, tin and iron can be almost entirely oxidized out of copper ores, while at the same time partially removing other impurities. Fluxing operations can be utilized to remove impurities such as lead, antimony and arsenic, but others such as nickel and bismuth can only be removed electrolytically. Electrolytic refining, which is a comparatively recent innovation in the processing of copper, is the only known, economic method for separating the precious metals, viz. gold, silver, etc. from copper, although it is not limited in its application to the removal of these metals.

Copper scrap, which is an important source of metallic copper, because of its collective heterogeneity contains many alloyed metals which must be removed before the copper can be re-alloyed with selected metals or used in the pure form. Since the largest percentage of commercially produced copper contains not more than about 0.3 weight percent impurities, excluding silver, it is apparent that impurity removal from either ore or scrap constitutes a major part of copper processing and, therefore, of copper costs. Copper used for electrical purposes, this constituting the largest single use of the metal, should not contain more than about 0.05 weight percent impurity content and preferably not more than about 0.01 weight percent to have acceptable electrical conductivity.

Patented June 1, 1965 whose impurity content is not more than about 0.1 weight percent.

Other objects and advantages of this invention will be in part obvious and in part explained by reference to the accompanying specification and drawings.

In the drawings:

FIG. 1 is a diagram showing the general relationship of the various steps of the copper process of this invention;

FIG. 2. is a somewhat diagrammatic flow sheet for the leaching operation used to obtain copper oxide;

FIG. 3 is a graph showing the degree of copper oxide reduction obtained at various temperatures using FeS as the reducing agent;

FIG. 4 is a graph showing the degree of copper oxide reduction obtained at variou temperatures using FeS as the reducing agent;

FIG. 5 is a graph showing the degree of copper oxide reduction obtained at various temperatures using FeS and Fes as the reducing agent and varying the copper oxide to iron sulfide ratio;

FIG. 6 is a graph showing the degree of copper oxide reduction obtained at various temperatures using ZnS as the reducing agent;

FIG. 7 is a graph showing the moles of copper reduced per mole of sulfide in reactants utilizing zinc sulfide reducing agent and in which other ingredients of the reaction mixture are altered;

FIG. 8 is a graph showing the degree of copper reduction obtained using elemental sulfur as the reducing agent;

FIG. 9 is a graph showing the effect of inert material in a reduction using elemental sulfur;

FIG. 10 is a graph showing the degree of reduction obtained using elemental sulfur to reduce cupric oxide to metallic copper;

FIG. 11 is a graph showing the effect of cupric sulfide in reducing cupric oxide to cuprous oxide at various temperatures;

FIG. 12 is a graph showing the degree of reduction obtained using cupric sulfide to reduce cuprous oxide to metallic copper;

FIG. 13 is a graph showing the degree of reduction obtained using cuprous sulfide to reduce cuprous oxide to metallic copper;

FIG. 14 is a graph showing the effect on the reduction rate of varying the amount of alkali metal chloride used in a reaction mixture containing cuprous sulfide and cuprous oxide;

FIG. 15 is a graph showing the degree of cuprous oxide reduction obtained using a mixed sulfide reducing agent and varying the amount of salt added to the reaction I mixture;

It is a principal object of this invention to provide a FIG. 16 is a graph showing the degree of reduction obtained when the oxygen content of the reaction mixture is varied;

FIG. 17 is a graph showing the degree of reduction obtained when varying amounts of cuprous chloride are added to the initial reaction mixture;

FIG. 18 is a graph showing the degree of reduction of cuprous oxide to metallic copper as a function of temperature in a reaction mixture utilizing a copper-sodium the chloridizing agents;

FIG. 19 is a graph showing the relative degrees of re duction obtained using sodium and potassium chloride as the chloridizing agent;

FIG. 20 is a graph showing the time required for completion of the reactions of FIG. 19 as a function of reaction temperature;

FIG. 21 is a graph showing the degree of reduction obtained with sodium or potassium chloride as the chloridizing agent when zinc oxide impurity is present;

FIG. 22 is a graph showing the'amountof copper reduced: using sodium or potassium chloride as the chloridizing agent when lead oxide impurity is present;

FIG. 23 is a somewhat schematic flow chart showing the arrangement for producing cuprous sulfide for use in the present process; and

FIG. 24 is a somewhat schematic flow chart showing the manner in which the reactants are reacted and the copper values obtained.

Basically, the process of this invention comprises reacting at elevated temperatures predetermined quantities of the active constituentsr (1) either cupric or cuprous oxide; (2) a sulfur-containing reducing agentwhich may be elemental sulfur or one or more of the sulfides of copper, iron, zinc, tin or lead; and (3) a chloridizing agent consisting of a chloride of either. an alkali or an alkaline earth metal, the chlorides of sodium, potassium or a mixture of the two being preferred.

Somewhat more broadly, the invention also includes an important leaching operation which is particularly efiicacious in obtaining copper oxide suitable for use in the process and also includes a means for preparing the sulfur containing reducing agent which is most effective;

The general nature and the inter-relationship of the steps used to obtain cuprous oxide or copper according to the invention may be seen best by referring to FIG. 1 of the drawings. Here block 20 represents the sulfur containing reducing agent which is fed into reactor 21. Blocks 22 and 23. represent sources of copper oxide, either cuprous'or cupric, andfan alkali metal and/or alkaline earth metal, respectively, which are also introduced into reactor 21. These basic ingredients are reacted in reactor 21 by heating to an elevated temperature, generally from 250 C. to about 750 C. depending upon the specific materials, added and result intended, to reduce the valence state of the copper oxide. Following reaction, a separation is effected, as at 24, to recover the desired-copper values at 25, while the salt by-products are recovered at 26.

The leaching operation of this invention is significant in that it provides a means for obtaining high purity cuprous and/or cupric oxide for subsequent purifying reduction. This leaching operation utilizes at least one important property of copper which heretofore has never been used, knowingly or unknowingly, in copper purification. This property is the inability for cuprous oxide (C11 or cupric oxide (CuO) precipitated from a liquid phase or from an aqueous solution, to accept more than the most minute traces of impurity in solution, regardless of the impurity concentration surrounding the precipitating oxide. The exclusion of impurities probably results from the fact that the copper oxides precipitate as precisely stoichiometric. compounds. Copper oxides crystallized or precipitated in the presence of impurities are therefore immediately capable of purification by selectively leaching out the impurities which mush-of necessity, have precipitated externally to the oxides.

I. LEACHIN G OF COPPER BEARING MATERIALS- COPPER OXIDE PREPARATION Although it was indicated above that the present inven tion is applicable to either copper-containing ores or metallic, scrap copper, it should be stated at the outset that ores to be economically feasible should, in most instances, be concentrated to raise the ratio of copper values to. waste. material. The process will work on ore having low copper content but the cost of chemicals for processing the .tonnages involved wouldbe generally prohibitive.

The basic leaching solution is composed of an aqueous solution of copper ammonium carbonate and ammonia, the constituents being present in the ionized forms: Cu(NH (4+5) (NH )aq.; 2(HCO At this point, it should be recognized that the ion formulas set forth represent idealized conditions and that in actuality they are probably much more complex. However, these idealized forms are felt to be adequate for purposes of explanation.

The independent variables of the leaching solution inelude:

[6:unknown excess quantity of dissolved ammonia] From this reaction equation, two factsmight be noted, first that the valence state of the copper in the copper amine complex is reduced from two to one, and secondly, that the dissolved ammonia and the acid carbonate ion remain unchanged. Thus, the feature responsible for the dissolution of copper is the ability of the cupric amine complex to incorporate an additional copper, thereby changing the amine from cupric to cuprous form.

After a small amount of copper ammonium carbonate is added'initia-lly to the leach solution, additional quantitles of the complex can be formed merely by adding ammonia ('NH;;) and carbon dioxide (CO to the S0111? tion before it is brought into contact with the source of copper. For optimum leaching capability, a ratio of NH to C0 of from about 34 to 1 can advantageously be used with a molar ratio of 3.4 to I felt to be best. The leaching capacity of the copper-amrnonium carbonate solutions is directly proportional to reagent concent-ration so that the use of any particular concentration is primarily a matter of choice based on sound eng neering and economic factors. The final variables which may be thought to be involved in the leaching operation are temperature and pressure. However, the effects of temperature and pressure on the leaching operation are not significant enough to include as determinative factors and can, therefore, be eliminated from consideration.

Referring to FIG. 2 of the drawings, the leaching operation is carried out in leach tank 30 by flowing the copper-ammonium carbonate leach solution through copper bearing material placed therein. The solution can be circulated repeatedly through tank 30 and connecting pipes 3-1, 32 and 33 by the pump 34 to increase the copper content of the solution to a level suitable for further processing.

A quantity of the solution is withdrawn by pump 34 and divided into approximately two equal quantities, half of the withdrawn liquid being filtered and heated to dissociate the cuprous amine ion andthe remaining half being oxidized to place all copper in the cupric condi Referring once again to FIG. 2, the pump 34 is used to remove a quantity of copper bearing-leach solution and half of the withdrawn amount is conducted, via pipe 32. and pipe 35 to pressure filter 40, which may be. of the usual packed-plate type. Filter 40 removes solids which are carried in the fluid stream. The solids are principally other metals and gangue which are not dissolved by the leaching solution but which. are in such a finely divided condition that they are carried by the fluid. The filtered solution now contains only copper, zinc and some extremely minor percentage of lead as metallic constituents. From filter 40, the non-filtered solution flows through pipe 41 into a tower 45 where it is heated to dissociate the copper amine complex. The

general reaction which occurs during heating may be expressed as follows:

[d unknown excess quantity of dissolved ammonia] It will be appreciated that other related reactions are simultaneously occurring with substances not present in stoichiometric proportions but that the preceding equation is representative of the group. Additionally, some cupric amine ions are undoubtedly present, so that some cupric oxide will be precipitated with the cuprous oxide. Generally speaking, cuprous oxide is the preferred product in this process due to the ultimate recovery of more metallic copper per unit of input material, as subsequently explained.

Heating of the filtered solution to dissociate copper amine complex may be accomplished by any suitable means. For example, tower 45 can be surrounded by a heating furnace capable of elevating the filtered solution to its boiling temperature. However, it has been diound that optimum conditions are achieved when the solution is heated by means of steam introduced directly into the solution within tower 45, as through the infusor 46 shown in FIG. 2. This type of heating is peculiarly adapted to use because it perm ts the precipitating cupric and cuprous oxides to settle to the bottom of tower 45 rather than plating-out on the walls of the tower as is the situation when external heaters are used.

The carbon dioxide and ammonia gases freed when the copper amine complex dissociates, rise to the top of tower 45 and are withdrawn and fed into a reflux condenser 47 through pipe 48. The condensate, mostly Water, can either be returned to tower 45 or discarded through pipe 49.

Removal of water in reflux condenser 46 results in a strong ammonium carbonate gas which is fed into total condenser 50 through pipe 51 for liquefaction. The liquid is then advanced through pipe 52 into a storage tank 55 where it is combined with the remaining onehalf of the withdrawn leach solution, following proccessing of the latter as subsequently described. Any gases which are not liquified in condenser 54) are fed into a water spray scrubber 56 and discharged.

The other half of the copper bearing solution, originally withdrawn from leach tank 30, is fed into a packed tower oxidizer 60 through the pipe 32 and mixed with air introduced via pipe 61 to oxidize the cuprous amine ion content to the cupric state. The oxidized copper ammonium carbonate solution is then withdrawn from the bottom of tower 60 and fed into the tank 55 through pipe 62 where it is mixed with the ammonium carbonate received from condenser 50. This solution, which is rich in ammonium carbonate, may then be used to replenish the supply used in dissolving copper in leach tank 30, since it will form more copper ammonium complex. Excess air present in oxidizer tower 60 is vented through pipe 63 to the water spray scrubber 56.

The entire purpose of the described leaching operation is to preferentially dissolve copper from copper bearing materials. However, as stated, most of any zinc present in the copper source material will be carried over with the copper along with small amounts of lead. Thus, the Cu O or CuO recovered will also contain some hydroxides and basic carbonates, with zinc as the primary-impurity. Purification of this mater al may now be accomplished by mixing it with cuprous chloride, forming a slurry and heating. When the slurry is heated to the boiling point, copper chloride hydrolyzes to copper oxide and hydrogen chloride and the acid generated dissolves the impurity precipitate. The net stoichiometric The reaction product may then be filtered and the copper oxides recovered, the zinc chloride being removed with the solution.

II. THE REDUCTION STAGE As stated earlier, the present process is carried out by combining, as active ingredients, cuprous and/or cupric oxide with a sulfur-containing reducing agent, including elemental sulfur and an alkali or alkaline earth metal chloride and reacting them at an elevated temperature in a protective atmosphere. Steam i a suitable protective atmosphere, although any which will not react with the copper being formed is also suitable. Specifically, elemental sulfur, copper sulfide, iron sulfide, zinc sulfide, tin sulfide or lead sulfide are suitable sulfur-containing materials which can be used to effect reduction of either cuprous or cupric oxide when combined with a suitable chloride and heated. It should, however, be pointed out that the sulfur-containing materials work with varying degrees of eificiency on the copper oxides. The most eflicient reducing agent has been found to be cuprous sulfide. This stage of the process, which is the crux of the invention, has therefore several, inter-related variables which are determinative if not of the theoretical operativeness of the process then of its posit-ion as a process which realistically advances the basic science and technology of copper metallurgy.

The principal composition variables are: (1) the choice of copper oxide, i.e. cuprous or cupric; (2) the selection of sulfur containing reducing agent; and (3) the selection of an alkali and/ or alkaline earth metal chloride. Additionally, variables such as the relative proportions of the various selected active constituents mentioned above, and the reaction times, temperatures and atmospheres must be considered in analyzing the reduction stage.

As set forth earlier, either cuprous or cupric oxide can be reduced according to this process, the monovalent cuprous oxide being reduced to the metallic state and the divalent cupric oxide being reduced either to the monovalent state or to the metallic state. Thus, the sole selection to be made at this point is whether it is desired to reduce cuprous or cupric oxide. In this connection, however, it should be indicated that the oxide product obtained from the ammoniacal leaching operation outlined earlier may result in the presence of both cuprous and enpric oxides in it final product, although the cupric oxide will normally be present in minor amounts only.

The list of metal sulfides which are thermodynamically capable of reducing copper oxides and which are available as low-cost, high-grade mineral concentrates are iron pyrite or marcasite (FeS reduced pyrite or marcasite FeS sphalerite or wurtzite (ZnS), and 'galena PbS). Metal sulfides of iron, copper, zinc, lead or tin can also be prepared by treating scrap metal with sulfur. The general form of the desired reaction for reducing cuprous oxide to metallic copper may be written:

(4) MS+2ACl+4Cu O A SO |-MCl +8Cu MS=metal sulfide ACl=alkali or alkaline earth metal chloride It will be recognized that the preceding equation is an idealized one and that in actuality side reactions will occur; for example:

If the metal sulfide (MS) is decomposed through reaction (5), then for complete reduction to occur in the system,

the final stage of the reduction process must decompose the metal oxide and convert it to chloride. The reaction may then .be written:

This reaction (7) has a positive free-energy change for some of the metals listed previously. For example, where the metal oxide (MO) is FeO, ZnO, SnO or PbO reacted at 1000 K., the free-energy changes for reaction (7) per mole of metal oxide are +145, +133, +5.1 and -5.0 kilocalories, respectively. In view of these energy changes, .complete reduction of copper oxide can be accomplished by the use of FeS, ZnS or SnS only if side reaction can be prevented from occurring to a noticeable extent or if one of the products of reaction (7) can be removed from the system as a gas or as a very insoluble compound.

Due to the uncertainties inherent in always accurately predicting the outcome of chemical reactions, samples using ;iron and zinc sulfides as the reducing agents were prepared and reacted to determine the eifectiveness of metal sulfides as reducing agents. In most of the following examples, sodium chloride was used as the salt but it should be recognized that the other alkali and/or alkaline earth metal chlorides can also be used and are contemplated Within the scope of the invention.

Example I.Reduction of cuprous' oxide to copper using iron pyrite Stoic-hiometric proportions .of reactants were prepared according to the reaction equation:

of Water being added to provide aprotective steam atmos-v phere in the ampules at elevated temperature. The ampules were preheated to 180 C'., and then placed in a salt bath for, varying periods of time ranging up to 1000 min utes. Salt bath temperatures of 660 and 700 C. were employed.

After being held at these temperatures for predetermined periods of time, the ampules were removed from the constant temperature salt bath, cooled and the material removed from the ampules, ground and analyzed for sulfate ion to determine the extent of the reduction reaction. It was found that the reaction was about 55 percent complete after 35 minutes at 700 C. and that the reaction moved no further toward completion as a result of holding the specimens at this temperature for much longer periods of time. By way of comparison, the specimens heated at 660 C. were only about 45 percent reacted after 70 minutes and ultimately reacted only to the extent of about 53 percent afterbeing held at temperature for-500 minutes.- The results can be seen by referring to FIG. 3 of the drawings wherecurve 65 shows the degree of reaction of the specimens heated at 700 C. and curve 66 the degree of reaction for the specimens heated at 660 C.

Example II.Reducti0n of cuprous oxide to copperusing ferrous sulfide (FeS In this instance,.iron pyritewas mixed with an excess over stoichiometric requirements of electrolytic iron and heated to 850 C. for one hour under a stagnant hydrogen atmosphere. The iron sulfide resulting from this treatment was separated from the excess iron and ballmilled under toluene to '100 mesh powder. A mixture was then prepared using stoichiometric requirements of 6 this sulfide, sodium chloride and cuprousoxide to correspond to the reactants for the reaction:

Cue-tenth mole of cuprous chloride. was added to encourage the formation of iron chloride and cuprous sulfide according to reaction (6) set forth earlier. This encouragement would be particularly useful during the first few minutes of reduction by minimizing other undesirable side reactions.

Specimens were prepared, heat treated, and analyzed as described in Example I. In this case, the reaction was about 45 percent'complete after 20 minutes at 700 C. and after about 60 minutes at 660 C. Holding for longer times resulted in progression of the reaction toward ultimate completion, the materials being about 50 percent reacted after about 500 minutes at 700 C. and after 900 minutes at 660 C. Curves 67 and 68 of FIG. 4 illustrate the nature of the reactions which occurred at 7 00 C. and 660 C., respectively. It is probable that the reaction does not proceed precisely as written in reaction (9) but rather that the ironroxidizes to the ferric state and forms an inert compound with copper oxide.

Example III.--Reducti0n 0f cuprous oxide to copper using ferrous sulfide (FeS Since iron sulfide produced by thermal decomposition of pyrite contains ferric iron which may have a detrimental efiect on the reduction process, the experiments of Example II were repeated using ferrous sulfide produced by reaction with sulfur, the reduced sulfide corresponding approximately to the formula Fes Reagent grade iron sulfide was pulverized by ballmilling under toluene and then mixtures of iron sulfide, sodium chloride, cuprous oxide and one-tenth mole cuprous chloride were prepared, placed in ampules, heat treated and analyzed in the same manner described in Example I. The reactants were mixed in the proportions shown in the following equation:

11 1/ 1 .05FeS +2NaCl+5Cu O and (12 V 1/1.05Fes, +2Nac1+4cu o were also prepared and treated in accordance with the process described in Example I. However, in these instances, thc first reaction (11) was only about percent complete whereas reaction (12) went only to about 55 percent completion. The curves 70, 71 and 72 of FIG. 5 of the drawings illustrate the nature and degree of completion of reactions (10), (11) and (12), respectively.

These experiments demonstrate that since iron in the system has a strong tendency to oxidize as ferric iron and precipitate as oxide, the iron sulfidereducing agent should be used in its most fully reduced state, i.e. the compound on the iron-rich side of the composition range.

Example I V.Reducti0n of cuprous oxide to copper using zinc sulfde (ZnS) Zinc sulfide ore concentrate. having the composition shown in the following Table I was ball-milled under .toluene to 325 mesh.

TABLE I.LOT ANALYSIS OF ZINC SULFIDE ORE CON- CENTRATE Element Percent I l Element Percent 59. 3 CaO 1. 2 31. 6 4

TABLE II.MOLAR RATIOS OF REACTANTS USED IN ZINC SODIUM OHLORIDE-CUPROUS OXIDE MIXTURES Mixture N0 1 2 3 4 5 Reactants (moles):

ZnS 1 1 1 1 1 NaCl 2 2 4 4 01.120 4 4 4 4 6 CuCl 2 1 Reaction products (moles):

SO 0. 002 0. 67 0. 67 0.67 1.00 Cu 9 9. 7

The reacted mixtures listed in Table H, following heat treatment at 660 or 700 C. were analyzed for sulfate ion and metallic copper. It will be noted from the reactants of mixture (1) and the resulting low sulfate content that sodium chloride, or more generally an alkali metal chloride or alkaline earth metal chloride, is an essential reactant in the reduction process. The low sulfate analysis from reacted mixture (2) indicates that incomplete reaction is present while the results of sulfate analyses on mixtures (3) and (4) show that excess metal chloride or cuprous chloride does not alter the reaction. The analysis of reacted mixture (5) shows that all of the sulfide ion is oxidized, oxygen of course being supplied by the cuprous oxide, in a mixture containing 4 and 6 moles, respectively, of sodium chloride and copper oxide. Mixtures (6) and (7) lowered the limits of sodium chloride and copper oxide to 2 and 55 moles, respectively, for essentially complete reduction.

The kinetics of sulfate formation, and therefore of the general reaction kinetics, were studied by heating several specimens of mixture (7) from Table II at 660 C. and 700 C. for predetermined times up to 1000 minutes. These specimens were closed in ampules and heated in the same manner as set forth in Example I preceding. It was found that the reaction rate was rather slow until the reaction is about 40 percent complete. Referring to curve 75 of FIG. 6 of the drawings, it will be seen that at 700 C. the reaction was about 40 percent complete in 10 minutes whereas at 660 0, about minutes was required for percent completion of the reaction process, as indicated by curve 76. The final 60 percent of the reaction occurred within a few minutes, the reaction being essentially complete after less than 15 minutes at 700 C. and after about 40 minutes at 660 C. Holding specimens for times past those at which the reaction was essentially complete resulted in some small fall-off of sulfate formation. Generally, for times exceeding 100 minutes, the reaction tended to reverse itself slightly so that after 1000 minutes at either 660 C. or 700 C. the reaction was only about 90 percent complete as compared to greater than 95 percent complete at the shorter times.

An unexpected finding resulting from analyses of reduced copper in the reacted specimens containing mixtures number (5) and (7) of Table II was the fact that 9 moles of copper were reduced. FIG. 7 of the drawings has plotted the moles of metallic copper per mole of sulfide ion oxidized in specimens treated at 660 C. and 700 C. for the indicated times. Specimens of mixture 5) are plotted as squares and identified by the numeral 77 and mixture (7) as circles identified by numeral 78. The basic reduction process acts through the oxidation of S to 8+ thereby releasing eight electrons which should reduce precisely eight moles of monovalent copper to metal. It is clear from FIG. 7, as was stated earlier, that nine moles of copper are reduced. Further, one specimen of mixture (5) held at 700 C. for 2000 minutes yielded 9.7 moles of copper.

Zinc sulfide appears to be a particularly attractive reducing agent for sulfide-salt reduction of copper oxides. One of its principal advantages lies in the fact that it is a natural resource which may be used as a reagent directly from the ore concentration mill. Iron sulfide, on the other SULFID E- hand, requires a thermal processing step to put it in its most fully reduced state before it can be used as a reagent. Additionally, the use of zinc sulfide as a reducing agent presents the possibility that silica in the ore concentrate can be rendered water soluble. Further, since the zinc is converted to a salt which is readily soluble in dilute acid, it is, in itself, far along the process for production of zinc metal by electrowinning from aqueous solution. In a process for refining low-grade scrap copper much of the zinc in brass can be recovered from liquors used in pickling the incoming scrap. This zinc can then be added to the reagent zinc. Liquors taken from the process of leaching reaction products away from the final reduced copper may be used for the scrap leaching treatment.

Example V.Reducti0n of cupric to cuprous oxide using elemental sulfur Stoichioimetric amounts of reactants were mixed both by wet and by dry methods according to the equation:

In the subsequent processing of these reactants, little, if any, difference was noted in the behavior of wet and dry mixed charges, so that this factor is not one which exerts any bearing on the outcome of the reduction. However, sodium chloride dissolves to an appreciable extent in an aqueous mixing medium and this dissolved salt may cause considerable difiiculty during the drying operation. For this reason, dry mixing methods were used when conducting further studies.

High temperature glass ampules holding 10 to 20 grams of powder were loosely filled after adding suflicien-t water to provide a percent steam atmosphere within the ampules on heating. The ampules were preheated to C. to drive off excess steam, then placed in a lead bath for times ranging from 1 to 1000 minutes at temperatures of 250 C. to 440 C. The ampules were closed during heat treatment by a one-way valve which allowed evolved gas to escape but prevented air from entering. After heat treatment, the ampules were analyzed for sulfate ion to determine the extent of the desired reaction.

FIG. 8 of the drawings shows in curves 80, 81 and 82, respectively, the percent of theoretical maximum sulfate ion found in specimens reacted at 250, 305 and 360 C. for the times indicated. The results from specimens heat areasaa treated at temperatures greater than 360 C. are not plotted since these specimens were found to be fullyreacted with heating times as short as 1 minute. It may easily be seen that specimens heated at 360 C. were more than 90 percent complete in 10 minutes (curve 82), and had become totally reacted at 100 minutes. By lowering the temperature to 305 C., the mixtures were 70 percent reacted in 10 minutes (curve 81),.and progressed steadily toward about 95 percent completion at the end of 1000 minutes. However, the specimens heated at 250 C. were approximately 85 percent complete in 10 minutes (curve 80), but actually became more incomplete as the heating was prolonged.

An experiment to measure the heating rate of the mixtures disclosed that specimens heated to 360 C. or lower temperature, reached the bath temperature in about 1 minute. When the bath temperature was greater than 360 C., the charge temperature rose to 600 C. within 2 seconds after insertion of the ampule in the bath. The heating effect was clearly the consequence of an extremely rapid reaction rate.

Since many of copper-bearing materials might contain a measurable proportion of inert material, tests were run to ascertain whether ornot the presence of large amounts of such inert material would in any way effect the reaction kinetics. Therefore, the isothermal reaction rates at temperatures close to the boiling point of sulfur were determined on specimens dilutedwith a large excess of silica powder to absorb the heat of reaction. Twelve moles of silica were added per mole of sulfur. The reaction rates of the diluted mixture at temperatures of 360, 380, 400, 420 and 440 C. are shown by curves 33-87 in FIG. 9 of the drawings. It may be noted that reduction at 360 C. proceeds at roughly the same rate here as did the material heated at 360 C. in FIG. 8 of the drawings. This demonstrates that the reaction is not clearly inhibited by inert matter separating particles of reactants. The curves of FIG. 9 of the drawings also show that the reaction rate continues to increase at temperatures above 440 C.

Example Vl.-Reductin of cupric oxide to copper using elemental sulfur The desired reduction equation in this case may be written;

A stoichiometric mixture was made in accordance with Equation 14, mixed, placed in ampules and heated in the same manner as the specimens in Example V. The results of sulfate analyses on reacted specimens are presented in FIG. 10 of the drawings, where it will be noted that about 65 to 75 percent of the sulfur oxidizes to the desired sulfate ion immediately but no further reaction occurs. FIG. 10 shows the temperaturesat which the specimens were reacted.

The experiments described in Examples V and VI show that sulfur is a highly efficient reducing agent for the reduction of cupric oxide to cuprous oxide and because of the rapid reaction rate and its highly exothermic nature, it should be possible to perform this reduction at very low cost in a commercial product. The sodium sulfate and copper chloride formed in the process can be separated by washing the mixture in sodium chloride brine and water. Copper couldthen be recovered from the brine solution by cementation on scrap iron. 7

Elemental sulfur is ofless value in performing reduction of copper oxide to the metallic state because of its tendency to divert by a side reaction to form copper sul- Thus, steps would have to be taken to preclude achieve essentially the same degree of completion.

12 Example VII.Rcducti0n 0 cupric to cuprous oxide using cupric sulfide The reaction for the reduction of cupric oxide to the cuprous state using cupric sulfide is as follows:

The cupric sulfide used in this example was prepared by hydrogemsulfide' precipitation from copper sulfate solution. The precipitate was washed to remove acid and then vacuum dried at room temperature. Stoichiometric amounts of reactants for Equation 15 were mixed in a dry ball-mill and specimens of the mixture prepared by loosely filling high temperature glass ampules containing a few drops of water.

These specimens were preheated to 180 C. to drive off excess steam, then heat treated in a lead bath in the same manner as was done with the samples in the previous examples. After heat treatment at 400, 500, 550 and 600 C., the reaction products were analyzed for sulfate ion to give a measure of the extent of the desired reaction. FIGURE 11 of the drawings shows that at temperatures of 500 C. and higher, the reaction is essentially complete within 5 minutes whereas at the lower temperatures, the reaction requires from 30 to .40 minutes to Curve illustrates the nature of the reaction at 400 C., curve 91 at 500 C., curve 92 at 550 C. and curve 93 at 600 C. It was found that the chloride ions in the reaction products were present as a mixture of cuprous and cupric chlorides rather than entirely as the cupric chloride as written in reaction (15).

Example VIII.--Reducti0n of cuprous oxide to metallic copper using cupric sulfide In this example, the theoretical, desired reaction is:

' (l6) CuS+2NaCl+4Cu O Na SO +2CuCl+7Cu Stoichiometric amounts of copper sulfide, as described in Example VII, cuprous oxide and sodium chloride were mixed by ball-milling under methyl alcohol. Dry ballmilling is difficult because of the sticky nature of the copper oxide. Ampules were filled with thernixture and heat treated at temperatures ranging from 600 to 700 C. after a pretreatment at 180 C. The specimens treated at 600 and 625 C. were loosely packed in the ampules while specimens treated at 630, 640, 660 and 700 C. were tightly packed. The density of these latter specimens is believed to be equal to the maximum density obtainable without closed die-pressing of the mixture. After heat treatment, the specimens were once again analyzed for sulfate ion to determine the extent of the desired reaction.

FIGURE 12 of the drawings clearly shows that'in all cases the reactionzstarts out fairly slow but accelerates rapidly once started. Specifically, curve 95 is for material heated at 600 C., curve 96 for material heated at 625 C., curve 97 for material heated at 630 C., curve '98 for material heated at 640 C., curve 99 for material heated at 660 C. and curve 100 for material heated at 700 C. The exceptions to slow initial reaction are those specimens heated at 700 C., where the reaction was immediate and essentially complete. The curves of the drawings clearly show that as the heat treating temperature is decreased, greater lengths of time are required to effect essentially complete reaction of the products.

Example IX.Reducti0n of cuprous oxide to metallic copper using cuprous sulfide I The following reaction is the one with which the present invention is most directly concerned since the success of the entire sequence of copper sulfide-copper oxide-sodium or potassium chloride reactions depend upon its ability to proceed to completion. This'reaction is:

To test the accuracy of the Equation 17, cuprous sulfide was prepared by passing hydrogen sulfide into a suspension of cuprous chloride and water, the precipitate was washed to remove acid and then vacuum dried. A stoichiometric mixture of this sulfide, cuprous oxide and sodium chloride, was then prepared by ball-milling under methyl alcohol. Specimens were prepared by loosely filling ampules containing a few drops of Water to provide a protective steam atmosphere, the ampules were preheated to 180 C. to drive ofi excess steam and then placed in a lead bath and heat treated at temperatures of 600 and 625 C. FIG. 13 of the drawings illustrates the extent of the desired reduction reaction at these temperatures, the samples heated at 625 C. (curve 105) becoming about 90 percent complete in slightly less than 200 minutes and more than 95 percent complete in slightly less than 500 minutes. As the figure shows, the rate of reaction of the material heated at 600 C. (curve 106) is slightly lower but moves to completion in about the same period of time as the material heated at 625 C.

It is apparent that in reactions of the type being dealt with in this invention that ancillary considerations must be undertaken with regard to amounts and types of materials being used, for example. These ancillary considerations deal mostly with locating sound operating parameters rather than with the basic effectiveness of the purifying operation, however.

For the experiments described following, cuprous sulfide was prepared by treating OFHC copper with sulfur at approximately 400 C. under a hydrogen-hydrogen sulfide atmosphere. The copper sulfide was crushed and ball-milled to 100 percent 375 mesh. A portion of the crushed product was mixed with sodium sulfide hydrate and heated to dehydrate and melt the mixture. The mixed sulfide was prepared to correspond to the composition (Cu Na S. After melting and casting, the sulfide was crushed and ball-milled under toluene to 100 percent 375 mesh and stored under toluene until needed. The mixed copper-sodium sulfide reducing agent was selected for use in the work described below for the purpose of eliminating copper chloride as a reaction parameter since none is produced as a by-product.

The mixture for this series of experiments was prepared by Weighing preselected amounts of the mixed coppersodium sulfide reducing agents, chemically pure sodium chloride, chemically pure cuprous chloride and cuprous oxide. The reagents were mixed in a mechanical blender using toluene as a dispersing fluid and the slurry filtered and the filter cake placed in small ampules for heat treatment. All ampules were preheated to 180 C. to drive off excess toluene and then heated at 700 C. for specified periods of time, the bath temperature being maintained at a selected temperature within 21 C. by a platinumrhodium thermocouple calibrated at the melting point of pure aluminum. After cooling, the heat treated specimens were analyzed for sulfate content to determine the extent of the reduction reaction.

Example X .-Efiect of excess sodium chloride FIG. 14 of the drawings shows the eifect of varying the sodium chloride content of a mixture containing 1 mole of cuprous sulfide and 4.1 moles of cuprous oxide. Curve 110 illustrates the degree of reduction occurring in specimens containing only the stoichiometric requirement, two moles of sodium chloride whereas curve 111 illustrates the degree 'of reduction occurring in specimens having one and two moles excess sodium chloride. The curves 110 and 111 indicate that excess sodium chloride'exerts very little effect on the early stages of the reduction process, but has a pronounced effect in depressing the kinetics of the latter portion of the reaction. Also, it may be noted that an excess of sodium chloride tends to drive the final equilibrium to the right, i.e. closer to 100 percent reduction.

TABLE III Composition (Gus-[5N 8.0.25)25+4.1 Ou;0+0.5 CuCl+4.0 NaOl.

In the reactions involving the mixed sulfide, the initial reaction rate is accelerated by excess sodium chloride; but the latter, more rapid stage of reduction is delayed so that the over-all elfect is to increase the time required to complete the reduction. The final equilibrium is shifted farthest toward completion with an excess of 0.5 to 0.7 mole of sodium chloride.

Example XI.Eflect of excess copper oxide The mixed reducing agent used here requires 4 moles of oxygen to be completely oxidized. FIG. 16 of the drawings shows the final portion of the reduction reaction in mixtures containing the stoichiometric oxygen requirement, 0.2 mole excess oxygen and 0.4 mole excess oxygen as cuprous oxide. The mixtures used are in Table IV below and are identified as to curve numbers.

TABLE IV Curve Composition As the curves 120-122 of FIG. 15 clearly indicate, a small excess of oxygen provides a small increase in the over-all reduction rate but does not alter the final equilibrium to a significant extent. Thus, so long as the cuprous oxide content is kept within reasonable limits of the stoichiometric proportions, the reaction will occur as desired.

Example XII.Efiect of additional copper chloride FIG. 17 of the drawings shows the effect of additional copper chloride on the reduction kinetics. The relationship between composition and curve is contained in Table V.

TABLE V Mixture Curve Composition I 123 (Cu0.15Na0.25)2S+1.6 Na01+4.4 CuEO. I 124 (Cuu.75Nau.25)z +1.6 Neel-F432 Cl1z0|0.5 CllCl. K 125 (Ouo.15Nau.2s)2S+1.6 NaCl+4A Cu2O+L0 CuCl.

The specimens containing no added copper chloride, curve 123, exhibited kinetics which differed markedly from the kinetics of reduction with pure copper sulfide. This difference could be attributed to the sodium sulfide component which may act to remove copper chloride from the system by the reaction:

is Na S+'2CuCl 2Cu S +2NaCl Mixture I (Table V)* contains just suflicient copper chloride to react with allof the sodium sulfide in thereducing agent. Here it may be seen that the reduction curve resembles fairlyclosely the reduction curve prolduced .by pure cuprou-s sulfide reducing agent. duction rate is still furtherincreased by a copperchlor-ide addition 0.5 mole greater than that required to react with The .re- 7 sodium sulfide, as evidenced by curve 125- of FIG. 16 of a the drawings (Mixture K, Table V).

Example XIIl.-Efiect of reaction temperature The studies of the chemical variables discussed earlier were conducted exclusively at 700 C., as already mentioned, to eliminate at least this one variable from the study of the reduction process. The findings obtained in a study of the effect of react-ion temperature made possible the formulation of a mixture of reactants which was a realistic practical compromise between rapid kinetics and a high percent reduction at equilibrium. The reactants studied here include: 1 mole of sulfide ion in a copper-rich, copper-sodium sulfide; sodium chloride equal to twice the normal fraction of copper in the sulfideplus 0.5 mole excessycopper oxides'to obtain 4.2 moles'of oxygen ion per mole of ion; and cuprouschloride equal to or greater than twice the mole fraction of sodium ion in the sulfide. A reaction mixture made up according to this formulation contained the molar proportions:

TABLE vL-MlXTURE L Composition (0110.15 Nae-2s): S|2.0 NaCl-l-4.2 Cu2O+0.5 CuGl. 0110.15 News); S+2.0 Nam-F42 Cuz0+O.5 CuCl. (C110.15.Nao.zs)2 S+2.0 NaCl-HZ CUzO-H).5 CuOl. (Cum New S+2.0 NaCl+4.2 CllzO-l-OI: CuCl.

Specimens were prepared from this mixture and sealed Thus far, while the importance of the chloridizing agent has been referred to, the important role which it plays has not been discussed in detail. Itiis well-known in the art that the process of chloridizing roasting entails salt.

mixed with metal sulfides followed by roasting of the mixture in an exces of air. The salt reacts with the sulfur trioxide produced by the combustion of the sulfur and' releases chlorine which converts them-etalvalues, notably copper and silver, to soluble chlorides. The key reaction occurring in this process may be described as:

where MCl is any alkali or alkaline earth metal chloride. The negative free-energy change for this reaction is so large that any alkali or alkaline earth metal chloride could be used with equal efiiciency.

The sulfide-salt reduction reaction of this invention, on the other hand, while using some of the same chemicals as are used in chloridizing roasting, requires no oxygen other than that combined with the metal oxide; This constraintbrings about a complete change in the nature of the reaction as compared to chloridizing roasting. Copper and silver oxides are reduced to metal with 'only a minor fraction of the whole converted tochloridea The free-energy change for the .reduction reactionis negative only for the cases where'an alkali or alkaline earth metal chloride is the chloridizing agent. 7 V

The alkali metal chloride serves a dual role in'the sulfide-salt reductionreaction, providingsodium'or potassium ions necessary for the formation of the sulfate salt and also providing chloride. ions to form the mixed copper-sodium or copper-potassium chloride flux in which the reduction reaction proceeds. There can be little doubt that both the reaction kinetics and the characteristics of the reaction product metal will differ, depending upon whethersodium chloride or potassium chloride, or one of the other alkali or alkaline earth metals, is used as the chloridizing agent, since both the free-energy change for the reaction and the fused salt flux are different in the two cases.

The principal reagents used in conducting the work described in the following examples consisted of chemically pure cuprous oxide, pure cuprous sulfide prepared by reacting OFHCcopper with sulfur at elevated temperature in a hydrogen-hydrogen sulfide atmosphere and chemically pure sodium and potassium chlorides. The copper sulfide and the chloride salts were each ball-milled 'under toluene to an estimated 200 mesh particle size.

Specimens were prepared by mixing weighed amounts of reagents in a blender using toluene as a dispersing fluid, the mixed specimen slurries then being filtered and dried and placed in ampules for heat treating experiments.

Heat treatment was accomplished by placing specimen ampules. in a molten aluminum bath controlled at a specified temperature il C, After heat treating, the specimens were analyzed for sulphate ion content to determine the extent of the reduction reaction. In some cases, 'the reduced copper was analyzed directly by leaching out the ionic salts and weighing the copper.

Example XI V.Reacti0n kinetics comparison Since both sodium chloride and potassium chloride are readily available and comparatively cheap materials, these two were investigated in some detail and a direct comparison of the properties of the two materials, both with regard to temperatures required for use and'the rate of reaction, were thoroughly investigated to find which would be preferable for use in the sulfide-salt reduction of copper oxide.

Two mixtures were prepared under identical conditions and were of the same molar composition, with the exception that one contained sodium chloride whereas the other contained potassium chloride. These mixtures are set forth in Table VII.

TABLE VII Mixtures Curve Reaction Composition temp, C.

M 135 700 C112S+2.5 KC1+4.2 C1120.

M 136 670 OIJ2S+2.5 KCl+4.2 CuiO. N 137 700 Cu2S+2.5 NaC1+42 CuiO.'

N 138 670 CllzS-l-2,5 NaGl+4.2 CuiO.

As Table VII shows, samples of the two mixtures were heat treated under identical conditions at 670 and 700 C. and the reaction product analyzed for sulfate ion content. The results show the percent completion of the reduction reaction, in terms of sulfate content, as a function of reaction time. It is clear that the reduction proceeds more rapidly in the specimen having potassium chloride as the chloridizing agent, curves and 137 of FIG. 19 being the potassium chloride curves at 700 and 670 C, respectively. Curves 136 and 138 show the reaction rates for the specimens using sodium chloride which were reacted at 700 and 670 C., respectively. FIG. 20 of the drawings shows the time required for completion of the reduction reaction as a function of reaction temperature.

It? intermediate of that of the mixtures containing sodium and potassium chloride.

11b. THE EFFECT OF IMPURITIES Thus far in the discussion it has been indicated that a copper purification process could be constructed using low grade scrap copper as input material, or concentrated copper ore, cupric ammonium carbonate leaching precipatation as a first purification step, and the sulfide-salt reduction as a final purification step. The principal metallic impurities remaining after the first purification step are zinc and lead. The present sulfide-salt reduction rocess otters two alternate reaction paths for a metallic oxide impurity. Specifically, it could be converted to a chloride dissolved in cuprous chloride or it could substitute for the alkali metal ion in forming the sulfate-salt. There are several possible ways in which zinc and lead could enter into the reaction. For example, zinc may remain inert as the oxide (ZnO) or be converted to the chloride (ZnCl or to the sulfate (ZnSO or it may be reduced to the form of metallic zinc. The lead, by way of comparison, may be changed to lead chloride (Pbclg), the sulfate (PbSG-Q, or to metallic lead, as well as remain unchanged in the oxide condition.

Considering in more detail some of the reactions which may occur in connection with zinc and lead oxide impurities in the copper oxide, the first case is that where zinc oxide remains inert during the reduction reaction. This reaction can be written:

the form of oxide or possibly combined with the sulfatesalt.

Considering now the instance where the zinc oxide is converted to chloride dissolved in cuprous chloride, the reaction is written:

The free-energy change for the reaction (21) is written at l000 K. and is +53 kcal. If KCl is substituted for NaCl, the free-energy formation is +49 kcal. In view of this, it is highly unlikely that any amount of zinc oxide will be converted 'to chloride. The degree of conversion to chloride was checked by making a mixture corresponding to the reactants of reaction (21) and heat treating specimens at 700 C. for and 900 minutes. Chemical analyses of the reaction products disclosed the presence of 8 percent of the total sulfate salt which could have formed. This amount could easily have been formed from the oxygen impurity in the cuprous chloride reagent. Even if it is assumed that all of the oxygen in the sulfate-salt came from zinc oxide, the results would show that cuprous chloride holds only 0.058 mole of Zinc chloride per mole of cuprous chloride. If the zinc oxide were introduced in the form of an impurity of cuprous oxide, a maximum of 1.5 percent zinc as oxide in copper oxide could be accomplished by solution in the chloride flux. 1

The conversion of zinc oxide to zinc sulfate was also checked and it was found that an appreciable amount of oxide was changed to the sulfate form. The optimum reaction for testing this reaction is written:

However, it is known that no reaction occurs if the alkali metal halide is omitted and, therefore, a compromise was made between reaction (22) and the mix- 18 ture of reactants, Cu S+2NaCl+4Cu O.' Thus, the reaction may now be written:

0.1 mole NaCl is estimated to be the amount remaining unreacted at equilibrium.

A mixture was prepared which contained molar proportions of the reactants listed in reaction (23) and a second mixture was prepared which was identical in every respect except that potassium chloride replaced sodium chloride. Specimens of the two mixtures were heated at 700 C., then analyzed for sulfate ion content. Referring to PEG. 21 of the drawings, curve shows the degree of sulfate formation at various time periods using the potassium chloride'salt. Curve 146, on the other hand, shows the degree of sulfate formation of the mixture using the sodium chloride. It can be seen from the list of reactants for reaction (23) that the system is deficient in either sodium or potassium ions. Had the zinc oxide not entered into the reaction, only 50 percent of the sulfide ion content would have been oxidized to sulfate. Since the reaction actually consumed 72 percent of the sulfide ions, it follows that 22 percent of the sulfate ions formed must have united with zinc ions to form zinc sulfate. Approximately 44 percent of the one-half mole of zinc oxide is converted to sulfate. If the zinc oxide is considered as an impurity in cuprous oxide, this would increase to 3.8 percent zinc in copper oxide. These results indicate that in a system deficient in alkali halide there is a strong tendency for formation of zinc sulfate, probably dissolved in sodium or potassium sulfate. It appears to matter little whether the alkali metal sulfate is potassium or sodium metal sulfate and in' fact it is quite probable that even in a system having a slight excess of alkali metal halide there will be a significant amount of zinc converted from oxide to sulfate.

Another possible reaction which must be considered with regard to zinc oxide impurity is that wherein the oxide is reduced to the metallic state. The reaction in this case can be written:

The free-energy change for this reaction at 1000 K. is +137 kcal. with sodium chloride as the chloridizing agent and +133 kcal. with potassium chlorideas the chloridizing agent. These figures indicate that zinc would not be reduced to the metallic state as a significant impurity in copper.

With regard to lead oxide as an impurity in copper oxide, the lead, as mentioned previously, might remain either as the oxide or be converted to a chloride or a sulfate or to the metallic condition. The situation in which lead oxide remains inert during the reduction reaction is analogous to that of zinc oxide expressed inreaction (20). That is, the amount of lead oxide remaining inert can only be determined by difference after the amount converted to chloride, sulfate or reduced to metal is known. e

The reaction. involving the conversion of'leadoxideto lead chloride may be written:

The'free-energy change for reaction (25) at 1000 K.

is -20 kcal. with NaCl as the chloridizing agent and -24 kcal. with KCl as the chloridizing agent. In view chloride.

Lead oxide is converted to lead sulfate combined with.

is or dissolved in sodium or potassium sulfate. The theoretical reaction for testing this reaction could be written:

The free-energy change for the preceding reaction is (-8) kcal. at 1000 K. so that one could expectthat conversion of lead oxide to lead sulfate would occur. In a system using potassium chloride as the chloridizing agent, the picture is somewhat more complex by virtue of. the existence of complex sulfate compounds, notably K2SO .PbSO and K2SO 2PbSO Since there is necessarily a negative free-energy change associated with the formation of a stable compound from its components it follows that the net decrease in free-energy for a mixture such as that in reaction (27) is greater when potas sium chloride is used in place of sodium chloride asthe chloridizing agent.

A mixture of reagents was made corresponding to those of reaction (27) and a similar mixture was also made utilizing potassium chloride rather than sodium chloride as the chloridizing'agent. Specimens of these two mixtures were heat treated at 700 C. for varying periods of time, then chemically analyzed to determine the percent completion of the reduction reaction. Because of the insoluble nature of lead sulfate, it was convenient in these experiments, to leach out the ionic salts and weigh the reduced copper directly. FIG. 22 of the drawings shows the percent completion of reaction (27) in specimens heat treated at 700 C. for the times indicated. Curve 147 illustrates the degree of completion obtained using sodium chloride at the chloridizing agent whereas curve 148 indicates the degree. .of completion obtained when potassium chloride was used. By comparing curves 147 and 148 with curves 133 and 136 of FIG; 18 of the drawings, it can be seen that the presence of lead oxide causes a marked acceleration of the reduction reaction.

The final reaction to be considered with regard to the lead oxide impurity is that wherein the oxide would be converted to metallic lead. This reaction is written:

The free-energy change for reaction (28') at 1000 K. is +12 kcal. when sodium chloride is used as the chloridizing agent and +8.5 kcal. when potassium chloride is used. In view of the positive free-energy change, it is not likely that lead oxide could be reduced to a lead-rich metallic phase. Since lead is soluble in copper to the extent-of only about 0.01 percent at 700 C., the amount which could be introduced as a soluble impurity would be very small. Further, since it has been shown that lead is readily removed from the reaction as a sulfate, the concentration of lead in'the regions where reduction is occurring should be lower than the average concentration in the' mixture.

The copper powder specimens resulting from the experiments described in connection with reaction (27) provided a convenient source of metal upon which to make ,a measurement of the purification ratio for lead in the sulfide-salt reduction reaction. Thepurification ratio is defined as copper/impurity output divided by copper/im-- purity input. The percentages of lead and copper in the input mixture are 14.4 weight percent lead and 85 .6' weight percent copper. The lead contents of the reduced copper powder specimens were determined by plating-out P130 from a perchloric acid solution containing the entire specimen. Lead contents of the eight copper specimens examined are listed in parts per million in Table VIII.

TABLE VIIL-LEAD CONTENT OF COPPER IRODUCED BY SULFIDESALT REDUCTION OF A LEAD-COPPER OXIDE MIXTURE Reduced copper- Purification Starting material, percent lead lead (in parts per ratio for lead million) It is expected that the maximum concentration of lead impurity which can be tolerated in the reduction process will be determined by the amount of lead sulfate which can be leached away from the copper after the reduction reaction is complete. Insoluble lead sulfate can be put into solution by leaching out with strong brine solutions of various kinds, although this technique is probably somewhat more expensive than a low-cost copper refining process should require. It-would probably be most economical to remove lead as completely as possible prior to the final reduction treatment so that the small amount remaining could then be easily put into solution by the same leaching treatment used to remove the alkali metal sulfate.

The development of a useful copper purification process utilizing sulfide-salt reduction requires the optimizing of two major objectives, viz. (l) the reduction of copper to the metallic state while leaving the impurities in an oxidizing state, either as oxide, chloride or sulfate, and (2) the reaction must be carried out in such a fashion that impurities are not physically entrapped in the copper particles. The reacted mixture must be capable of undergoing leaching treatments of relatively simple nature which issolves the impurities while leaving the copper values unattached. Thealkali metal sulfate plays a vital role in entrapment of impurities by keeping copper particles separated from one anotherduring the reduction treatment.

The spacing of copper particles in the reacted mixture will be' greater with potassium sulfate because of the larger molar volume of the potassium salt. Molar volumes are 52.7 cm. /g.-mole for Na SO and 65.5 cm. /g.- mole for K 80 The basic mixture, i.e.

when fully reacted would contain 34.4 volume percent copper with sodium chloride as the chloridizing agent and 31.9 volume percent copper with potassium chloride as the chloridizing agent. While this difference is seemingly small, it conceivably could have a large effect on the amount of entrapped impurity. The difference in specific volumes of sulfate salt favors the use of, potassium chloride.

.With regard to the relevance of the melting points of the principal ionic salts, i.e. sodium and potassium salts or mixtures thereof, the reaction does not become rapid until an appreciable quantity of copper-alkali metal chloride flux has been formed. This stage occurs earlier with potassium chloride than with sodium chloride as the chloridizing agent because of the lower melting point of the former; 776 C. for potassium chloride and 801 C. for sodium chloride. Also, the melting point of the sulfate salt is important. In the event that severe overheating of the reacted material should occur, the sulfate would melt and the copper particles would condense into a sponge, trapping-impurities. The higher melting point of potassium sulfate (1076. C.) as compared to sodium sulfate (884 C.) would favor the use of the potassium salt to ensure against difiiculti'es from sulfate melting.

Although potassium chloride is a costlier reagent than sodium chloride of equivalent impurity, the additional 23. cost of potassium chloride is more than offset by the greater salability of potassium sulfate. Sodium sulfate has a marginal value at best while the potassium salts are necessary constituents of fertilizers.

The comparison of sodium and potassium chloride with respect to their efiect on the kinetics of the sulfidesalt reduction process disclosed that the potassium salt permits the use of a reaction temperature which is C. lower than is possible if sodium salt is used. A comparison of the efiect of impurities lead to the important discovery that impurity elements lead and zinc precipitate to a considerable extent in the sulfate salt. Thus, the reduction. process proceeds concurrently with an internal purification process which buries the impurity in the oxidation product. All of the elements which form stable sulfates, i.e. bismuth, cadmium, cobalt, manganese, nickel, lead and zinc will behave similarly.

HI. PREPARATION OF COPPER SULFIDE REDUCING AGENT FIG. 23 of t e drawings illustrates the manner in which cuprous sulfide, which is the preferred reducing agent for reasons already covered, (C11 8) can be produced. Numeral 1553 indicates a cupola into which is charged coke (carbon), copper, sulfur and sodium sulfide. Air is blown upwardly through the charge by means of blower 151 which is connected to the tuyres 15.2. The furnace is heated with natural gas until the coke ignites and then combustion of coke supplies the heat. The sulfur reacts with copper to form cupric sulfide which in the presence of carbon is reduced to cuprous sulfide. The carbon also reduces the sodium sulfide to sodium sulfate. These two reactions, which occur simultaneously within cupola 15% are expressed in the following two reaction formulas:

The mixed sulfides, that is cuprous sulfide and sodium sulfate, have a lower melting point than cuprous sulfide along. This molten sulfide is tapped from cupola 156 into a ladle 155 and then cast into pig molds 156 for solidification. After solidification is complete, they are fed into a small jaw crusher 157 and then into a hammer mill 15$ for crushing into smaller particle size. The resulting product is then suitable for use in the basic reaction of this invention to produce metallic copper or copper oxide having a low impurity content. The crushed powder may be leached with water to remove sodium sulfide, out this operation is not essential to the process.

The reduction reaction and subsequent removal of copper powder from the reacted products can best be unerstood by reference to FIG. 24 of the drawings; The cuprous and cupric oxides received from the ammonium carbonate leach step, or from any other suitable source, is combined with the cuprous sulfide and with a selected alkali or alkaline earth metal chloride. The preferred ingredients are, as previously explained, cuprous oxide, cuprous sulfide and potassium chloride or sodium chloride. The mixed copper oxide-copper sulfides should be analyzed to check the oxygen tosulfur ratio of the sulfur compounds. This ratio should be maintained at approximately 4.1:1 1-1 percent for maximum operation, although a ratio within the range of from 4:1 to :1 is acceptable. Fine adjustments in the sulfur-oxygen ratio could be made by first having a small amount of copper chloride in the mixture and then adding ammonium hydroxide or sodium sulfide solution to the. mixture to in: crease either the oxide or sulfide ion content of the precipitate. This mixture is combined withsodium or potassium chloride as in the blender 160 shown in FIG. 23 and poured into trays 161 for firing at the preselected temperature within furnace 162. The firing temperature'may range from 250 to as high as 775 C. or higher, depending upon the particular materials charged. The upper temperature limit is limited only by such considerations as the possible vaporization of reactants or reacted products or by the arising of back reactions which lower the extent to which the reactions proceed toward completion. As a general matter, lower temperatures are desired as a matter of economics. The copper sulfide-copper oxide-potassium, sodium-chloride materials mentioned generally benefit from operating temperatures of from 600 to 775 C., this therefore being the preferred reacting temperature for these materials. Following reaction, the materials are sent on through a cooling zone 163 and then either stored until use is desired or crushed and sent on for further processing to recover the copper values.

The reduction process which occurs in furnace 162 leads to a sponge which contains copper, sodium sulfate, and copper chloride. The techniques which might be used to separate the mixed powders include dry air elutriation, water elutriation or heavy media separation. Dry air elutriation offers the best arrangement for selectively removing copper from the mixture, the advantage accruing by virtue of the fact that copper powder is ductile while all other constituents are relatively brittle. It is, therefore, possible to grind the reacted mixture to the point where the ionic salt particles are smaller than the copper particles, thus permitting separation through utilization of the different settling velocities of small particles 'in a gaseous medium.

A second characteristic of the mixture which facilitates dry separation is of course the great difference between the densities of metallic copper and the ionic salts with which it is mixed. Copper has a density of 8.92 grams per cc. whilesodiurn sulfate and copper chloride have densities of 2.698 and 3.53 grams per cc., respectively. Also, impurity particles, silica etc., have densities which generally are lower than that of copper.

Referring once again to FIG. 24, the reaction products from cooling zone 163 of furnace 162 are introduced into a jet mill 165 through the inlet opening 166. Simultaneously, air is introduced to the interior of mill 165 through the air inlet openings 167, this air escaping via the outlet tube res. The air exiting through .tube 168 carries with it the particulate material so that upon entering cyclone separator 169 the comparatively heavy copper powder settles to the bottom and can be periodically withdrawn. The air continues its flow and leaves cyclone 169 through the tube 17%, carrying the reaction salts with it into a second cyclone 171. It is in this cyclone that the sodium sulfate and any copper chloride salts are trapped for recovery. The remainder of the air and any ultra-fine particles continue on to a bag filter 172.

Although it would be highly desirable to completely separate the salt from the metallic copper, it is unlikely that air separation can effect such a complete removal.

Therefore, the copper powder is placed within a washing tank which contains a weakly acidified waterisolution, a 10 percent hydrochloric acid solution being suitable. After the powder has been thoroughly agitated in bath 175, the solution is removed and replaced with deionized water to remove any traces of acid which may cling to the particles. This solution is then pumped into a vacuum filter $.76 where the water is removed so that the copper powder can be moved into an annealing furnace 177 for drying and agglomeration. Since the surfaces of the particles probably will pick up some minor amounts of oxygen, the annealing should be conducted in a slightly reducing atmosphere containing hydrogen .or crack-ed ammonia to reduce any copper oxide which may have formed. The copper which is obtained by virtue of eifecting'the process described will compare favorably with electrolytic copper. v g

Although the present invention has been described in connection with preferred embodiments, it is to be understood that modifications and variations may be re? sorted to without departing from the spirit and scope of the invention, as those skilled in the art will readily under 23 7 stand. Suchmodifications and variations are considered to be within the purview and scope of the invention and the appended claims.

What I claim as new and desire to secure by Letters Patent of the United States is:

1. A process for producing cuprous oxide or copper metal by reducing, a copper oxid from the group consisting of cupric oxide, cuprous oxide and mixtures thereof to a lower valence state, the steps of the process comprising: providing a mixture in which the active constituents are; (a) a reducing agent selected from the group consisting of sulfur, copper sulfide, iron sulfide, zinc sulfide, tin sulfide and lead sulfide, (b) said copper oxide, and (c) a chloridizing agent selected from the group consisting of alkali metal chlorides and alkaline earth metal chlorides; and heating the mixture in a protective atmosphere to an elevated temperature sufiicient to chemically react the active constituents and achieve reduction in the valence state of said copper oxide.

2. A process for producing cuprous oxide or copper metal by reducing a copper oxide from the group consisting of cupric oxide, cuprous oxide and mixtures thereof to a lower valence state, the steps of the process comprising: providing a mixture in which the active constituents are; (a) a reducing agent selected from the group consisting of sulfur, copper sulfide, iron sulfide, zinc sulfide, tin sulfide and lead sulfide, (b) said copper oxide, and (c) a chloridizing agent selected from the group consisting of alkali metal chlorides and alkaline earth metal chlorides; and heating the mixture in a protective atmosphere to an elevated temperature for a time sufficient to react the active constituents and reduce at least 90 percent of said copper oxide to a lower valence state.

3. A process for producing cuprous oxide or copper metal by reducing a copper oxide from the group consisting of cupric oxide, cuprous oxide and mixtures thereof to a lower valence state, the steps of the process comprising: providing a mixture in which the active constituents are; (a) a reducing agent selected from the group consisting of sulfur, copper sulfide, iron sulfide, zinc sulfide, tin sulfide and lead sulfide, (b) said copper oxide, and (c) a chloridizing agent selected from the group consisting of alkali metal chlorides and alkaline earth metal chlorides; and heating the mixture in a protective atmosphere to a temperature of not lower than about 250 C., reacting the active constituents and reducing the valence of the copper in said copper oxide. 7

4. A process for reducing cupric oxide to cuprous oxide comprising: providing a mixture in which the active constituents are; (a) sulfur, (b) cupric oxide, and (c) a chloridizing agent selected from the group consisting of alkali metal chlorides and alkaline earth metal chlorides;

and heating the mixture in a protective atmosphere to a temperature not lower than about 250 C. for a time sufficient to react the active constituents and reduce the cupric oxide to cuprous oxide.

5. A process for producing metallic copper from a copper oxide selected from the group consisting of cupric oxide, cuprous oxide and mixtures thereof compn'singi providing a mixture in which the active constituents are; (a) copper sulfide, (b) said copper oxide, and (c) a chloridizing agent selected from the group consisting of alkali metal chlorides and alkaline earth metal chlorides and heating the mixture in a protective atmosphere to a temperature of not less than about 600 C. for a time sufficient to react the active constituents and reduce said copper oxide to metallic copper. V

6. A process for producing metallic copper from cuprousox-ide comprising: providing a mixture in which the active constituents are; (a) cuprous sulfide, (b) cuprous oxide, and (c) a chloridizingagent selected from the group consisting of alkali metal chlorides. and alkaline earth. metal chlorides and heating the mixture in a protective atmosphere to a temperature of not less than about 600 C. for a time sufiicient to react the active constituents and reduce the cuprous oxide to metallic copper.

7. A process for producing metallic copper from cuprous oxide as defined in claim 6 wherein said mixture contains the active constituents; (a) cuprous sulfide, (b) cuprous oxide, and (c) sodium chloride.

8. A process for producing metallic copper from cuprous oxide as defined in claim 6 wherein said mixture contains .the active constituents; (a) cuprous sulfide, (b) cuprous oxide, and (c) potassium chloride.

9. 'A process for producing metallic copper from cuprous oxide comprising: providing a mixture in which the active constituents are; (a) cuprous sulfide, (b) cuprous oxide, and (c) potassium chloride, the ratio of Cu O to Cu S being from about 4:1 to 4.511 and the ratio of chloride in the potassium chloride to sulfur in the cuprous sulfide being from 2:1 to 2.5:l, heating the mixture in a protective atmosphere to a temperature of not less than about 600 C. for a time suflicient to react the active constituents and reduce the cuprous oxide to metallic copper.

10. A process as defined in claim 9 wherein the mix ture is heated to a temperature of from about 700 C. to 775 C.

11. A process for producing metallic copper from cuprous oxide comprising, preparing a mixture consisting essentially of cuprous sulfide, cuprous oxide and sodium chloride and heating the mixture in a protective atmos phere to a temperature of not less than about 600 C. to cause reaction between the constituents thereof, the cuprous sulfide, cuprous oxide and sodium chloride being present in proportions such that the reaction proceeds substantially according to the formula:

12. A process for producing metallic copper from cuprous oxide comprising, preparing a mixture consisting essentially of cuprous sulfide, cuprous oxide and potassium chloride and heating the mixture in a protective atmosphere to a temperature of not less than about 600 C. to cause reaction between the constituents thereof, the cuprous sulfide, cuprous oxide and potassium chloride being present in proportions such that the reaction proceeds substantially according to the formula:

13. A process for producing metallic copper from cuprous oxide comprising: providing a mixture in which the active constituents are: (a) cuprous sulfide, (b) cuprous oxide, and (c) sodium chloride, the ratio of Cu O to Cu S being from about 4:1 to 45:1 and the ratio of chlorine in the sodium chloride to sulfur in the cuprous sulfide being from 2:1 to 2.5:1, heating the mixture in a protective atmosphere to a temperature of not less than about 600 C. for a time sufiicient to react the active constituents and reduce the cuprous oxide to metallic copper.

14. A process as defined in claim 13 wherein the mixtureis heated to a temperature of from about 700 C. to 775 C.

. 15. A process for producing metallic copper compris ing: dissolving copper values from a source of copper with an ammonium leach solution to form a cuprous amine, heatingthe cuprous amine to. effect dissociation thereof and precipitate cuprous oxide, preparing a mixture by combining the cuprous oxide with a reducingagent selected from the group consisting of sulfur, copper sulfide, iron sulfide, zinc sulfide, tin sulfide and lead sulfide and a chloridizing agent selected from the group consisting of alkali metal chlorides and alkaline earth metal chlorides, and heating the mixture in a protective atmosphere to an elevated temperature suflicient .to chemically react the mixture and reduce the cuprous oxide to metallic copper.

(References on following page) References Cited by the Examiner UNITED STATES PATENTS 26 OTHER REFERENCES Hot'man et al.: Metallurgy of Copper, 2nd Edition, McGraw-I-Iill, N.Y., 1924, pp. 69-74.

5--72 ggi g Jacobson: Encyclopedia of Chemical Reactlon-s, Volume j 5 III, Reinhold Publishing Corp., 1949, pp. 329, 337, 352, 1,103,258 7/14 Braekelsburg 75-111 5 60 42 4 7 d4 2,385,066 9/45 Du Rose 23-447 3 9, 3 5, 2 ran Newton et al.: Metallurgy of Copper, John WIIey & FOREIGN PATENTS Sons, N.Y., 1942, pp. 65-71. 514,098 6 55 C d 0 BENJAMIN HENKIN, Primary Examiner. 634,006 1/62 Canada. DAVID L, RECK, Examiner.

Claims (1)

1. A PROCESS FOR PRODUCING CUPROUS OXIDE OR COPPER METAL BY REDUCING A COPPER OXIDE FROM THE GROUP CONSISTING OF CUPRIC OXIDE, CUPROUS OXIDE AND MIXTURES THEREOF TO A LOWER VELENCE STATE, THE STEPS OF THE PROCESS COMPRISING: PROVIDING A MIXTURE IN WHICH THE ACTIVE CONSTITUENTS ARE, (A) A REDUCING AGENT SELECTED FROM THE GROUP CONSISTING OF SULFUR, COPPER SULFIDE, IRON SULFIDE, ZINC SULFIDE, TIN DULFIDE AND LEAD SULFIDE, (B) DAID COPPER OXIDE, AND (C) A CHLORIDIZING AGENT SELECTED FROM THE GROUP CONSISTING OF ALKALI METAL CHLORIDES AND ALKALINE EARTH METAL CHLORIDES, AND HEATING THE MIXTURE IN A PROTECTIVE ATMOSPHERE TO AN ELEVATED TEMPERATURE SUFFICIENT T CHEMICALLY REACT THE ACTIVE CONSTITUENTS AND ACHIEVE REDUCTION IN THE VALENCE STATE OF SAID COPPER OXIDE.

Images