US2060074A - Copper refining apparatus - Google Patents

Description

Nov. 10, 1936. R. P. HEUER COPPER REFINING APPARATUS Filed April 15, 1932 2 Sheets-Sheet 1 o R E 2 an VAIQIJAED 5511x950 NI BZIIUVAIEUMBL lumen/far J.

Nov. 10, 1936. R. P. HEUER COPPER REFINING APPARATUS Filed April 15, 1932 2 Sheets-Sheetfi Patented N v. 10,. 1936; v 2,060,074

UNITED STATES PATENT OFFICE corraa REFINING APPARATUS Russell Pearce Heuer, Bryn Mawr, Pa., assignor to The American Metal Company, Limited, New York, N. Y., a corporation of New York Application April 15, 1932, Serial No. 605,389 In Canada March 23, 1931 9 Claims. (CI. 22-57) My invention relates to apparatus by which ing. The process introduces many impurities the refining of copper may be carried out. into the metal which seriously impair the grade The present invention is a continuation in part of the products. of my copending application, Serial No. 459,300, Copper cathodes commonly carry copper sulfor Copper refining methods, apparatus and prodphate on the surface, or occluded between copper 5 uct, filed June 5, 1930. crystals. This sulphur, as well as sulphur from A purpose of my invention is to produce copper the gases of combustion, forms as an impurity substantially free from cuprous oxide and at the in the molten copper. Oxygen is also present in same time having so few gas cavities as not to the bath as dissolved cuprous oxide. The blow- 10 interfere with the mechanical working of the ing, while removing part of the sulphur, increases 10 copper. the oxygen content of the bath and saturates the A further purpose is to reduce by means of carbath with nitrogen and other gases from the bon the content of cuprous oxide to a point much atmosphere. below that reached under the present practice. The primary purpose of the poling operation A further purpose is to crowd carbon into the is to remove the excess of oxygen, present largely 15 body of a substantially sulphur-free copper melt in the form of cuprous oxide. In doing so, howin order to reduce the cuprous oxide content to a ever, the bath takes up hydrogen. carbon monoxminimum. ide, carbon dioxide and water vapor from the A further purpose is to prepare copper castings organic matter.

by roasting solid copper cathodes to substantially At times during the poling operation the bath 20 eliminate the sulphur, melting the cathodes, reis protected by charcoal or coke floating upon the ducing any cuprous oxide present in the molten surface. By virtue of its position and its small copper by contact with carbon, avoiding blowing quantity as compared with the bath. the amount and poling, and obtaining an oxide-free copper of reduction possible from carbon used in this having no dissolved gaseous impurities except manner is negligible. :5 carbon monoxide. V Previous observers have noticed that, when the A further purpose is to prepare copper for castpoling operation is carried too far, the copper is ing by eliminating sulphur contamination'from valueless because of the large numbers of gas the solid copper by reduction with hydrogen, and cavities. Asstated by H. O. Hofman (Metallurgy removing any dissolved hydrogen by melting in of Copper, McGraw-Hill Book 00., Inc., New 30 an oxidizing atmosphere. York, 1914, page 384) It is important that some A further purpose is to obtain copper castings C1120 remain in the copper, as the dissolving free from any excess of carbon dioxide, water power of copper for gas is greatly weakened by vapor or hydrogen and consequently without gas its presence. The general feeling has been that cavities due to the liberation of these impurities in some way the presence of CllaO decreases the 35 during the solidification period of the copper. solubility of the copper to gases, and consequently A further purpose is to reduce carbon dioxide to avoids the danger of gas cavities. Another excarbon monoxide. in the copper preliminary to planation which has been put forward is that the casting. copper would become more highly saturated with 40 Further purposeswill appear in the speciilca reducing gases as poling proceeded further, and 40 tion and in the claims. the solubility of the copper for the gases would In the drawings, Figure 1 is a top plan view decrease on cooling, causing them to separate as of an arrangement of furnaces by which my gas cavities during solidification. It is commonmethods may be desirably performed. ly considered bad practice to reduce the content Figure 2 is an enlarged vertical section of the or C1130 below 0.3%. 45 carb n reducing furnace n in Figure The reduction of cuprous oxide under the con- Figures 3 to 5 are diagrammatic views useful ditions of poling takes place according to the folin explaining my invention. lowing reactions: pail; the drawings like numerals refer to like cuzo+c:zcu+co I (1) 50 The common practice is to melt copper cath- CusO-{-CO=2Cu-}-(JC (2) odes in large intermittent reverberatory furnaces. In such furnaces, the conventional mode of op- Cu20+m 2cu+mo (3) eration is as follows: (1) charging; (2) melting; I will consider the above reactions in terms of 5 (3) blowing; (4) skimming; (5) poling; (6) castthe thermodynamic activities of the reacting com- 6 ponents and o! the standard free energy changes for each reaction where the activities of the reacting components are reierred to arbitrarily chosen standard states for the reacting components in which their activities are unity. (See Gilbert Newton Lewis and Merle Randall, Thermodynamics and Free Energy oi Chemical Substances, McGraw-Hill Book Co., Inc., New York, first edition, 1923. cha ters XXII md XXIV, and pages 190, 255, and 2 3.) Thus for the reaction Since at constant temperature the logarithm becomes constant, the expression of Equation (5) is frequently simplified under the name so! the law or mass action". When this is done, the pressures of gases and the molalities of dilute solutions are ordinarily substituted tor the activities.

Applying Equation (5) to Reaction (1). and using the same nomenclature, we get:

(LCu) (A.C0) Am-RT i.cu,o(s..c)

(A.Cu) =activity of copper (A.CO)=activity or carbon monoxide (A.Cu-.-O) =activity of cuprous oxide (A.C) =activity of carbon In the above reaction equilibrium is never reached 7 during poling. It is only attained when the cuprous oxide is practically all removed.

Applying the general form of Equation (5) to Reaction (2), and using the previous nomenclature, we obtain:

RT a.c u;o a.co

(A.CO:) =activity of carbon dioxide It is probable that the bath will become saturated with carbon dioxide during the poling operation,

so that (A.CO2) will be constant. Also, (A.Cu) is constant as the bath is practically pure copper.

- Then Equation (7) becomes:

where K is a constant for the particular conditions. At constant temperature this expression becomes:

(A.Cu:0) (A.CO) :K1 (9) Ki=proper constant Equation (9) has been plotted in Figure 3, using (A.CuzO) values as the abscissae and (A.C0) values as the ordinates.

Using the same symbols, the expression for Reaction (3) is:

(A.Cu)'(A.H O) "APPRT (A.Cu,O A .I-l,)

(A.Hz0) =activity of water (A.H1)=activity of hydrogen the concentration. Therefore, we may observe Since the activities. of copper and of water are constant, this equation becomes, at constant temperature:

(some) (Alb) =K: (u) Ka=proper constant Equation (11) has been plotted in Figure 4. using (A.Cuz0) values as the abscissae and (A111) values as the ordinates.

The activity of any component is a function at the relation between the concentration 0! carbon monoxide .or hydrogen and that oi the cuprous oxide by the use 0! the curves shown in Figures 3 and 4.

For example, with a high cuprous oxide concentration, corresponding to the point a, the concentrations of hydrogen and carbon monoxide are low. As the concentration of cuprous oxide falls. as shown progressively at points b, c, and d. the concentrations 0! hydrogen and carbon monoxide remain substantially the same.

When the cuprous oxide content reaches a low enough value, as that indicated by e or I, at the knees of the curves, the concentrations oi hydrogen and carbon monoxide rise. When the concentration of cuprous oxide is reduced progressively below that represented by the point I. a rapid rise in concentration of hydrogen or carbon monoxide occurs, entirely out of proportion to the decrease in the concentration of cuprous oxide.

In the practical operation of the refining turnace, poling is stopped at about the point a on the respective curves of Figures 3 and 4, representing a cuprous oxide content 01' about 0.3%. Previous observers have considered that it is necessary to stop at this point, either because the cuprous oxide reduces the solubility of gases in the copper, as suggested by Holman, or because the copper will become more highly saturated with reducing gases as the poling proceeds i'urther, and thee'e reducing gases will separate out on cooling. Both oi! these explanations have been found by me to be erroneous.

As previously stated it is not the separation of a carbon monoxide and hydrogen because of lower solubility in solid than in liquid copper, or any other reason, which causes the bulk of the gas cavities, but rather the reactions during solidification of the dissolved carbon monoxide and hydrogen with any remaining cuprous oxide to produce carbon dioxide and water vapor with which the copper has been saturated already.

By reference to Reactions (1), (2) and (3) it will be seen that Reaction (1) can hardly take place during the solidification oi the casting, because no carbon is present at that time. However, Reactions (2) and (3) could occur, since both carbon monoxide and hydrogen are present.

During the solidification oi copper containing less than 3.45% of cuprous oxide, crystals of pure copper will first separate out, thus concentrating the cuprous oxide in the supernatant liquid along the line DE as shown in the conventional temperature-composition curve for the system onprous oxide-copper, reprodpced in Figure 5. (Bee Holman, Metallurgy oi 'Copper, McGraw-Hill Book Co., Inc., New York, 1914, page 13.)

At the eutectic temperature (E in Figure 5), 1065 C., the liquid will contain 3.45% of cuprous oxide. This increase in the concentration oi cuprous oxide during solidification causes the Reactions (2) and (3) to proceed rapidly, yielding carbon dioxide and water vapor.

Thus it will be seen that overpoling and exees- II sivegas cavities are due just as much to the presence of cuprous oxide in the copper as they are to the presence of reducing gases, such .as hydrogen and carbon monoxide. I propose therefore to eliminate cuprous oxide, thus avoiding the injury done to the copper from overpoling and excessive gas release, and by producing copper free from cuprous oxide, I obtain better physical properties and higher electrical conductivity in the product.

The solubility of hydrogen in molten copper is considerably greater than its solubility in solid copper at the freezing temperature. On the other hand, there is substantially no change in the solubility of carbon monoxide when copper solidifies. Therefore, forgetting for the moment the question of reactions with cuprous oxide, carbon monoxide is a. much less undesirable ingredient in molten copper than hydrogen. I avoid the presence of excessive quantities of hydrogen or remove it before casting, whether or not cuprous oxide is present. Carbon monoxide,.however, is not particularly harmful from the standpoint of change ofsolubility on solidification, nor of formation by any reaction possible during solidification, so that I may even saturate the bath with carbon monoxide as later explained.

Poling is necessitated by blowing, and blowing is required in order to remove sulphur. To eliminate both blowing and poling I propose to eliminate sulphur from the cathodes prior to melting them, rather than to attempt to oxidize sulphur out of the molten bath. Having eliminated sulphur, I melt the cathodes under conditions which prevent any absorption of sulphur.

My low sulphur bath does not require blowing, and therefore does not have the high oxygen content due to blowing, and the saturation with other atmospheric gases for which it is responsible. The elimination of poling enables me to avoid the attendant contamination with hydrogen, carbon dioxide and'water vapor.

Hydrogen, carbon dioxide and water vapor tend to produce gas as previously explained, but carbon monoxide does not of itself act in this way because it is not formed by any of the reactions which occur during cooling, and its solubility does not greatly decrease when the copper solidifies.

As a further precaution against the possibility of reactions occurring between hydrogen or car.-'

bon monoxide and cuprous oxide during cooling, I reduce the cuprous oxide in the resultant metal much below any content which has been deemed possible heretofore. The reduction of cuprous oxide until only a trace is present is best performed by a large exposed surface of solid carbon in contact with the bath.

Having substantially eliminated the cuprous oxide, I go further, and reduce the carbon dioxide in the molten copper to carbon monoxide, using the carbon monoxide formed to protect the copper against further cuprous oxide formation.

In the drawings I show apparatus suitable for producing my refined copper, but do not intend to indicate that this is by any means the only apparatus which might be employed. The roasting or calcining operation to remove sulphur is best performed in the continuous flue of a fur nace as shown in my United States Patent No.

' 1,914,716, for Copper melting furnace, granted The furnace A of Figure 1 is of the general type shown in my patent, No. 1,914,716, above referred to. Cold cathodes are charged at 8 by any suitable pusher Ill on to the-continuous hearth I i. operating in the flue l2. The-hearth is moved by a motor i3 and any suitable connections. In the illustration shown the piles of cathodes ll are carried clockwise to the discharge point" I 5, where they are picked up by the fork l6 and carried into the furnace l1 and there dropped. The charging fork I6 is desirably moved in and out through a door l8. I In the furnace II, the bath l9, desirably covered by a suitable protective slag, is heated by products of combustion from burners at 20. I prefer to use oil burners, but it will be understood that any suitable means of heating may be employed. From the furnace H the products of combustion pass through the neck 2|, through which the cathodes are charged, and around the flue in a counter-clockwise direction, finally rising through the stack 22. During normal operation of the furnace, the products of combustion are prevented from taking a clockwise path by the wall 23.

The hearth ll in the fiue l2 may be continuous or may consist of a plurality of cars separately moved about the fiue. The continuous mode of operation is much superior, and therefore I indicate the hearth as extending around the fiue and being turned by any suitable driving means '3.

The preheating chamber l2 contains a non-- Lower temperatures will suffice, although operation is slower in that case. Higher temperatures will work satisfactorily provided care is taken not to melt the cathodes in the flue. Melting here is objectionable for various reasons, metallurgically because molten copper will dissolve the copper oxide and copper sulphide formed, reducing the thermodynamic activity and preventing a quantitative elimination of sulphur.

While I consider it preferable to desulphurize in a roasting or calcining flue, I may also desulphurize solid copper in furnaces free from combustion gases, as, for example, muflies or electric furnaces. In furnaces of this type I may desulphurize by reducing with hydrogen, so that copper sulphate, copper sulphide and similar sulphur compounds will give up their sulphur in the form of hydrogen sulphide. Here also, the presence of molten copper is undesirable for the reasons above stated.

Where reduction with hydrogen is employed, I will preferably melt the copper under slightly oxidizing furnace conditions to remove dissolved hydrogen, which might otherwise produce gas cavities in the castings.

The preheating operation in a furnace of the type of furnace A should last about one hour, although the time taken by a given pile of cathodes in traversing the flue varies with the size of the furnace, and the type of cathode.

In order to melt the cathodes, I prefer to plunge them into a bath of molten copper, preferably suitably protected from contamination with sulphur from the flue gases by a suitable slag. The

slag used may be the normal slag forming on the bath from oxidation of the copper, or it may be a special slag as desired.

It will be understood that where copper is already free from sulphur, it may be charged into furnace II without preliminary calcining to eliminate sulphur. However, as the movement through the flue serves to preheat the copper charges, thus increasing the economy of operation, I prefer even in that case to use a furnace of the type of furnace A. It would of course be possible to preliminarily desulphurize and intermediately cool the cathodes before charging them into furnace II.

Having melted the sulphur-free copper, I next proceed to reduce by carbon any oxygen present, thus avoiding the possibility of reactions with reducing gases which may be dissolved in the copper. No blowing is necessary since the sulphur is already low, and, since no blowing is required, a

the copper need not be poled.

In order to reduce the cuprous oxide, I pour the metal from the pool I! of furnace A into furnace B. Furnace B is preferably free from gases of combustion, and is heated electrically or by any other suitable means. In my figure I show electrodes ill passing through the wall of the reducing furnace, and sealed, to prevent entrance of air, by collars 26 and 21. The electrodes are connected to a suitable current source, not shown. Copper from the bath I9 is preferably charged into the inlet opening 28 of furnace B through a closed trough 29 heated by a burner Ill. Slag is retained in the furnace I! by a gate ll.

The furnace B is filled, both above and below the level of the copper, with coke or other suitable carbonaceous material. For this purpose coke is forced in through the openings 32 and 33 near the top, by means of plungers 32 and 33'. The heating in the furnace is partly due to the electrical resistance of the coke surrounding the electrodes, since the electrodes will not usually be immersed in the bath. The copper bath 3! may be drawn oil as desired through the tap opening 35 and the pouring spout 36 at the end of the furnace.

In ordinary operation the level of the pouring spout is slightly above that of the copper bath 34, so that the furnace must be rotated slightly in order to hour from the spout. For convenience in rotation I preferably make my furnace B circular in cross section. For rotating the furnace I provide suitable supporting rollers 31 engaging bands 38, and also a gear driving band 38 engaged by a rack 40 which is moved back and forth by a screw I through a non-rotatable nut 42, secured to the rack 40. A work hole 43 is provided in the top of the furnace.

The tilting mechanism for the furnace may be used to agitate the copper in contact with the carbon, if desired, but I do not consider agitation necessary to accomplish the reduction of cuprous oxide. 4

The reactions taking place in the reducing furnace during the first stage of reduction are as follows:

Carbon monoxide produced in Reaction (12) will dissolve in the copper and tend to saturate the copper. Reaction (13) will remove some carbon monoxide from the copper, and carbon dioxide formed from Reaction (13) will tend to saturate the copper.

The carbon monoxide is not objectionable because it is not produced by the reaction between any remaining cuprous oxide and the reducing gases during solidification, nor does its solubility change markedly during solidification. But I wish to have the copper bath unsaturated with carbon dioxide to allow for the possible formation of carbon dioxide during the cooling period by oxidation of carbon monoxide. Therefore, I continue the contact between the carbon and the copper until all of the cuprous oxide is reduced. When this has taken place the carbon will reduce carbon dioxide as follows:

c+oo==2co (14) In order to reduce carbon dioxide to carbon monoxide, I prolong the deoxidation after the cuprous oxide is effectively eliminated. Any substantial excess of carbon monoxide above that required to saturate the bath will pass oil at this stage, and the carbon monoxide passing oi! will protect the copper from contamination with other gases. Thus I will obtain copper substantially free from cuprous oxide and containing practically no dissolved gas other than carbon monoxide.

The carbon monoxide dissolved in the copper has a "buffer" action in protecting the copper from increase in cuprous oxide content. This carbon monoxide, produced by Reaction (14), is formed from carbon dioxide which was formerly dissolved in the copper, so that after Reaction (14) is complete, the carbon dioxide will be removed from the copper, and carbon monoxide produced instead. Notwithstanding that two volumes of carbon monoxide are produced from one volume of carbon dioxide, most of the carbon monoxide resulting from Reaction (14) will remain dissolved in the copper because carbon monoxide is quite soluble in molten copper, while carbon dioxide is only sparingly soluble under the same conditions.

Now, as previously pointed out, it is possible that an effective quantity of oxygen will be introduced into the copper before the copper solidifies, either due to leakage of air into the copper deoxidation furnace or due to contamination with air at the pouring spout or in the mold during casting. By its "buffer" action, the carbon monoxide will care for a slight contamination with oxygen by reacting with the oxygen to form carbon dioxide. The "bufler action of carbon monoxide is in addition to the ability of carbon monoxide to react with existing cuprous oxide in the supernatant liquid during solidification. Since the copper has been deliberately unsaturated with carbon dioxide. the carbon dioxide resulting from the "buffer" action of carbon monoxide will not oversaturate the copper, and so will not produce gas cavities. It is of course obvious that protection from oxygen is highly important because the quantity of oxygen which may be cared for by the "bullet" action of the carbon monoxide is limited.

The reduced copper from furnace B will ordinarily be poured into molds. In Figure l I illustrate diagrammatically a conventional casting wheel having supporting structure II and molds 45, adapted to be moved in either direction to a position registering with the pouring spout 8., when copper may be poured through openings 46 into the molds. In order to minimize contact between the air and the oxide-free copper during casting, and decrease the absorption of oxygen, I pour preferably into vertical molds. By

this method I eliminate the set" or oxidized copper surface obtained on horizontal castings. For best results I also surround the stream of pouring copper with a nonoxidizing gas, preferably carbon monoxide, since the copper under ordinary conditions will already be saturated with carbon monoxide,. to prevent the absorption of atmospheric gases, especially oxygen.

As explained elsewhere in the specification, hydrogen and water are harmful and should not be allowed to contaminate the deoxidized copper. I illustrate at 1 conventional means for surrounding the stream of pouring copper with an atmosphere of carbon monoxide. To avoid excessive introduction of air I provide a shield 48 covering the top of the pouring spout and the mold, desirably supported from the reduction furnace at 49.

Thus it will be seen that I produce my novel copper product by the following steps, desirably in succession: (l) removing sulphur from the cathodes; (2) melting; (3) reducing cuprous oxide; (4) reducing carbon dioxide; (5) casting. The first step might be eliminated if sulphur-free material could be obtained, and of course the sulphur might less desirably be removed after melting. The last step, while offering a desirable combination in the practice of my invention, could be dispensed with if it were desired to transfer the copper to a storage furnace, for example.

Independently of the role of cuprous oxide, the presence of hydrogen in the copper bath is undesirable because of its tendency to form gas cavities due to its decrease in solubility during solidification. Green poles previously used to reduce copper oxide of course introduce a large quantity of hydrogen into the bath in the form of hydrocarbons. By avoiding the use of material rich in hydrogen, as by employing charcoal or coke, I avoid introducing hydrogen into the bath.

As a less desirable alternative, I may effect the early stage of the deoxidation by means of material containing hydrogen, such as green poles, completing the removal of cuprous oxide and reducing carbon dioxide to carbon monoxide by material free from substances rich in hydrogen. As noted in Figure 4, the hydrogen contamination during the early stage of deoxidation is relatively small, andsome hydrogen introduced during the early stage will be mechanically removed during the later stage of deoxidation.

It will be understood that my invention could be carried out intermittently in one furnace, where the calcining, melting, and reducing operations would proceed. This, however, would be slow and less desirable than the method of operation using two or more furnaces, either continuously or intermittently.

It will be evident that the elimination of sulphur is necessary in those cases only where the raw copper material has an excessive sulphur content-which represents the normal condition--; also that the elimination of blowing and poling desirably keeps down the amount of cuprous oxide and carbon dioxide present and avoids the necessity for excessive reduction by carbon to take care of this condition, although blowing and poling can be practiced if desired, providing the poling is stopped before excessive removal of the cuprous oxide is accomplished; that the final reduction of cuprous oxide and subsequently of carbon dioxide below the saturation point of carbon dioxide is highly essential and is secured most quickly and most effectively by immersion of the carbon within the molten copper bath; and that casting within vertical molds reduces the subsequent air contamination with its increase of cuprous oxide and carbon dioxide within the solid copper.

In view of my invention and disclosure variations and modifications to meet individual whim or particular need will doubtless become evident to others skilled in the art,'to obtain part or all of the benefits of my invention without copying the structure shown, and I, therefore, claim all such in so far as they fall within the reasonable spirit and scope of my invention.

Having thus described my invention, what I claim as new and desire to secure by Letters Patent is:

1. In apparatus for treating copper, a fuel-fired copper melting furnace, a deoxidizing furnace free from fuel combustion filled with carbon which is in direct contact with the copper and reduces cuprous oxide and changes carbon dioxide to carbon monoxide, a mold adapted to receive the copper from the deoxidizing furnace and means for displacing the air of the mold by a harmless gas during casting to protect the copper from the atmosphere.

2. In apparatus for treating copper, a preheating and roasting flue adapted to eliminate sulphur'from copper charges, a fuel-fired copper melting furnace, a deoxidizing furnace free from fuel combustion filled with carbon which is in direct contact with the copper and reduces cuprous oxide and changes carbon dioxide to carbon monoxide, a mold for forming the copper into a casting and means for displacing the air of the mold by a harmless gas during casting to protect the copper from the atmosphere.

3. In apparatus for treating copper, a fuel-fired copper melting furnace having the gases of combustion in contact with the bath of molten copper, a deoxidizing furnace free from oxidizing and contaminating gases, and containing carbon which is in direct contact with the molten copper and reduces cuprous oxide and changes carbon dioxide to carbon monoxide, a mold for receiving deoxidized copper from the deoxidizing furnace and a body of harmless gas containing carbon monoxide within the mold during casting to protect the deoxidized copper from the atmosphere.

4. In apparatus for treating copper, a preheating and roasting flue for eliminating sulphur from solid copper charges, a fuel-fired copper melting furnace in which the molten copper is exposed to the gases of combustion, a deoxidizing furnace free from contaminating combustion gases, and containing carbon which is in direct contact with the molten copper and reduces cuprous oxide and changes carbon dioxide to carbon monoxide, a mold for receiving deoxidized copper from the deoxidizing furnace and a body of harmless gas containing carbon monoxide within the mold during casting to protect the deoxidized copper from the atmosphere.

5. In apparatus for treating copper, a preheating and roasting flue for eliminating sulphur from solid copper charges, a fuel-fired copper melting furnace, an electrically heated deoxidizing furnace containing carbon for removing oxygen from the molten copper, a mold for receiving deoxidized copper, means for introducing carbon monoxide about the copper while it is pouring into the mold to protect the copper from the atmosphere and a shield above the mold for conlining the carbon monoxide and excluding the atmosphere.

6. In combination. a deoxidising furnace containing molten copper and having a roof, means for forcing carbon into the deoxidizing furnace and maintaining a deep bed of carbon against the roof, which opposes the tendency of the carbon to float, a mold. a conduit communicating molten copper and having a roof, means for forcing carbon into the furnace against the floating tendency of the carbon so as to maintain a substantial bed of carbon in the furnace preming against the roof and submerged beneath the copper and means for agitating the furnace to move the molten copper with respect to the carbon during deoxidation;

9. In combination, a fuel-fired copper melting furnace, a deoxidation furnace substantially filled m with carbon submerged beneath the copper, means whereby the deoxidation furnace is agitated to improve the contact between the copper and the carbon, a mold and a conduit communicating with the interior of the mold for introll ducing a permanent gas of protecting character to prevent contamination of the copper by the atmosphere.

RUSSELL PEARCE mm.

CERTIFICATE OF CORRECTION.

Patent No. 2,060,074. November 10 l RUSSELL PEARCE HEUER It is hereby certified that error appears in the printed specification 1 the above numbered patent requiring correction as follows: Page 5, second column, line 72, claim 5, for the word "means" read a conduit communicatin with the interior of the mold; and that the said Letters Patent should be read with this correction therein that the same may conform to the record of the case in the Patent Office.

Signed and sealed this 2nd day of March, A. D. 1937.

Henry Van Arsdale (Seal) Acting Commissioner of Patents.

Images