US2193482A - Process for the production of alloys of beryllium and copper - Google Patents

Description

Patented Mar. 12, 1940- UNITED STATES amuse raooass roa 'rna monoc'rrou for, armors or Baarmm AND oorrna Andrew J. Gahagan, New York, N. Y., assignor to The Beryllium Corporation, New York, N. Y.,

a corporation of Delaware No Drawing. Application November 14, 1938,

Serial No. 2 0,340

3 Claims. '(01. 75-40) This invention-relates to a process for the isolation of alloys of beryllium and-copper. More-- over, this invention afiords a process for the continuous manufacture of commercial beryllilj -um-copper alloys directly from the oxide, such continuous manufacture having hitherto been impossible in actual practice. a

The fact that aluminum could be reduced from its oxide by carbon in the presence of cop- "per and other metals, thereby yielding alloys of these metals, has been known ever since the work of Cowles in 1884. The process was technically quite feasible on a large scale, but it failed to survive commercially because the cheap "electrolytic method for manufacturing pure aluminum from bauxite made the co-reduction procedure unnecessary. In view of the resemblance of beryllium to aluminum, it seemed, from the very first, reasonable to assume that this procedure of co-reduction would succeed just as well with beryllium. The first technical work along this line was that of Lebeau, who, in 189?, formed small amounts of alloys of beryllium by co-reduction of the oxides, or by reduc- 5- tion of beryllium oxide in the presence of the metals, using the electric are as a source of heat."

In connection with the work of Lebeau, see Comptes rendus, vol. 125, l897pages 1172-1174. At about the same time Liebmann (see German Patent No. 94,507) discussed the same general type of process. Later developments and disclosures were made in this field by Rohn (Urs Patents Nos. 2,025,614 and 2,025,616) and Burgess (U. S. Patent No. 1,905,340). As late (Wiss. Veroilnt. Siemens Konzern, 11, 88-92).

While the foregoing accomplished the production of some alloys, especially in small scale or oxide of the two metals employed in the process.

Furthermore, the prior workwas hampered by 50 impurity of'the alloys produced. Still further,

at least some of the prior processes involved even though Rohn obviated the presence of the mixed the use of special equipment. Thus,

oxide compound in beryllium-copper, this was 5 only accomplished by the use of a special atas 1932, scientific work was carried on by Kroll mosphere, thereby displacing the normal carbon Q equilibrium. 4

The present invention has as oneof its primary objects to overcome difficulties of the type mentioned and to provide a procedure which 6 would be at once simple and eflective, yet re-. quire nothing but relatively standard equipment. This I have accomplished as set forth below, my new procedure yielding an alloy free I from the above-mentioned red double oxide, and 10 flowing readily outof the furnace, Thereby I have made possible a continuous process, in itself a sharp diiferentiation from previously known procedures, wherein it was necessary to stop a batch run, dig out the arc-treated maten'al, and separate the metal from the residue.

Not only does the beryllium-copper alloy formed under normal operating conditions contain what appears to be a skeleton of double oxides, as indicated by Lebeau, but I have-also observed that much of itremains physically entrapped in the larger mass of beryllium oxidecarbon mixture. This latter condition interferes, certainly because of physical factors, and probably also by virtue of mass action, with the 95 I further continuance of the reduction. I have found, as one part of my invention, that if after the usual preliminary heating of a mixture of carbon, beryllium oxide, and finely divided copper (or an oxide of copper with sufflcient excess of carbon to reduce it to metal) I subsequently add more metal and permit it to melt and flow down through the reaction mass, the metal washes down the beryllium-copper alloy already formed and thereby permits the reaction to be 5 renewed. Simultaneously, the alloy is freed, by this molten metal washing action, from the skeleton of double oxide, which latter is then left, in open, highly reactive condition, unprotectedby the alloy, for furthefre'duction. 40

This washing down procedure, of course, di-

vlutes the alloy somewhat, but not merely to the extent that might be assumed from earlier workers' results. For example, alloys of about 5% beryllium can be obtained in small quantity by remelting away 'fromtheoxide skeleton in accordancewith the procedure ,flrst indicated by Lebeau; my procedure permits manufacture; on

a much larger scale, of an alloy running regularly El /2% in beryllium, and occasionally considerably higher. Yields are high; practically all the beryllium values are recovered, either as metal or as re-usable oxide. It will be understood that the addition of massive copper to wash down the alloy formed can be made with copper metal of any reasonable macroscopic particle size. There .is no advantage to using finely divided metal at this point of operation, but it will be understood, in the claims, that massive metal refers to any copper suitable for washing down the alloy already formed.

I have found that good results are obtainable from the use of massive metal in the form of chunks or pieces of copper, such chunks or pieces being introduced into the furnace preferably alternately with charges of a mixture of carbon, finely divided copper and beryllium oxide.

Thus, after the preliminary heating of a batch of carbon, finely divided copper and beryllium oxide inthe furnace, some massive copper is added so as to initiate the washing down action and operation of the furnace is carried out continuously, with periodic additions, desirably alternated, of a mixture of carbon, finely divided copper and beryllium oxide, on the one hand and massive copper, on the other hand. I have found that in a furnace appropriate for carrying out the present process, introduction of a mixture of carbon, copper and beryllium oxide may suitably be effected at intervals of the order bf'about one hour. Massive copper may suitably be added at similar alternate intervals, although when using chunk copper, the addition thereof is effected at more frequent intervals; for instance several times between each addition of the mixture of carbon, copper and beryllium oxide.

The intervals of introduction of both the mixture and the massive copper will, of course, vary in accordance with a number of conditions, including furnace size, size of particles of massive copper used, etc., although as-a general guide it may be stated that the massive copper should be added in a manner which will practically continuously maintain a reserve of copper present adequate to perform the washing down process above referred to.

. It is obvious that not only beryllium oxide as such, but any of the compounds which will break down under heat and/or reduction to yield the oxide, can be equally well used. For example, the hydroxide, the basic carbonate, the sulfate, the nitrate, and organic derivatives, are all equally well usable for the purpose of my, invention. In general, the oxide has been found most economical to use, and the phrasing of this patent is therefore in terms of the oxide itself.- In the claims, it will be understood that beryllium oxide refers to ore and to the compound itself whether added as such to the mixture or whether formed by decomposition or dissociation in situ.

While I have found it most convenient and economical to use finely divided copper in elementary form for-the preliminary reduction, it is an obvious alternative, which I have found equally satisfactory, to use a mixture of finely divided oxide of copper, with suflicient excess carbon or other reducing material to yield copper on heating. Where the term copper is used in the claims, it should theerfore be understood that such, copper may be vadded'either in elementary form or in the form of an' oxide or similarly reducible comsmall volume, in which the full temperature'of the are (about 4000. C.) holds sway; Beyond this, the following various zones may be considered a to exist.

(1) The first of these zones may be viewed as represented by the sphere wherein occurs the greatest temperature drop, from about 4000 C. to about 2300 C. In this sphere, no liquid metal can exist, since both beryllium -and copper are vaporized above 2300 C.

(2) This volume will, in turn, be followed by a shell representing the 2300 C. to 1900" C. drop.

In this shell, carbon is soluble in copper (a factor I of great importance in the reduction of the beryllium oxide), while any beryllium which may form will have a very high vapor pressure, even in the presence of alloying copper.

(3) In the next shell, from 1900" C. to 1500 C., molten copper will no longer dissolve carbon, but the other conditions are essentially the same Copper is in the fused state, while beryllium vapor pressure is still quite high.

(4) In the zone which follows, from 1500 C. to 1285 C., there is both copper and beryllium still in a state of free fusion, with comparatively low vapor pressure on the part of the beryllium. This represents the condition of optimum alloying if reduction problems were not inherent in the situation, and nothing else were involved but the union of free elementary beryllium and copper.

(5) In the next shell, from 1285 C. to 1083 C., the former being the melting point of beryllium and the latter the melting point of copper, one of the components, beryllium, is no longer in the molten state. i

(6) Below 1083 C., and reaching to about 850 C., we have a zone of primarily theoretical although not devoid of practical interest, wherein beryllium-copper alloy remains liquid, freezing at some point within that temperature range, corresponding to its composition. Below 850 C., the melting point of the copper-beryllium eutectic, all possible components are of course" solid, and neither reaction nor alloying can take place.

Experiment has shown that conditions for reduction of the beryllium oxide, for alloying of the nascent metal, and for maintenance of the alloy undecomposed, are all 'diiferentand that, therefore, furnace construction must be such as to give each of these essential factors opportunity to play its normal part. It appears, from the unpublished work of others, that while they did maintain conditions satisfactory for the reduction of the beryllium oxide, and perhaps also for' alloying action, these conditions were such as to decompose the alloy almost as rapidly as it was being formed. As a consequence,- recoveries were always so low as to make the are reduction method impractical.

Whether carbon can reduce beryllium oxide alone in the 4000" C.-2 300 C. core is purely of theorctical interest; evl' .if reduced to metallic state, the beryllium sos-"f med, lacking copper as a collecting agent (the'zone is above the normal boiling point of copper),-would immediately be volatilized away'or, more likely, converted to car-' bide. What little did escape. carbonization, or reoxidationby reversal of the reaction:

BeO+C=Be+CO would condense in the cooler portions of the furnace in the form of thin metallic films on the. reaction mass, or be volatilized completely out of the furnace.

Wherereduction reaction does occur is pri- 2,198,482 marily in the 2300 C.-1900 C. shell. It is here that carbon dissolves in copper, and it is this copper-carbon alloy that appearsto react best with the beryllium oxide, dissolving the beryllium immediately as formed.

But this zone, in which beryllium can best be divorced from its oxygen, is not conductive tomaintenance of the alloy just'formed. The high vapor pressure of elementary beryllium in that temperature range willgradually cause that element to volatilize away from the alloy; once alone as an element, either as vapor or as condensate and unprotected by a preponderanceof copper around it, the beryllium is rapidly car-- bonized or oxidized.

It is consequently essential that there be, best completely surrounding the 2300 0.4900 0.

zone, or certainly aboveand below it in the furcalled for in my invention are; There must be sufficient heat conductivity from the bottom of the crucible or container, or, more accurately,

. from the liquid alloy already formed, to maintain its temperature below the point where the beryllium vapor pressure becomes strong. enough to aiford material loss. This may be done by There should be a water cooling, for example.

" sufllcient "head of reaction material above the arc zone to permit the existence of at least a 1500 C.-1285 C. shell, without that shell overheating, by conduction and by contact with the heated carbon monoxide resulting from the chemical reduction, to beyond the 1500 C. where beryllium begins to vaporize awayappreciably.

It is also very desirable that the 1500 C.-1285 C, zone extend around the. sides of the are as well-as around the top and bottom: to'accomplish'this, it is necessary that the arc furnace be of sufllcient diameter to permit conduction and convection elects to counteract the heat of the arc, and reduce the temperature on the outside periphery "to .1500 C.-l285 C. desired for alloy maintenance. And finally, it is preferable although not essential to have a diameter which provides a surrounding zone of from 1285 C. to 850 C. I -In practiced find it, preferable to maintain, above the arc, suillcient charge sothat there is a full temperature drop to as low as 850 C. This insures that all beryllium vapor rising out of the more heated zones will be trapped,'either physicentral point equivalent to an arc. The condi-- tions surrounding arc operations willnot be signiiicantly changed iasregard temperature zones even if the mode of heating is thus changed. To

i indicate the type of heating wherein the'maximum' tempera'ture is at a focal point, around or,

. 3 adjacent which there are zones of lesser temperature, -I have chosen to term "focal heating,"

and in the claims this term is to be understood as covering arc heating, inductive heating, etc.

Where the arc is used, I prefer to use a three I phase furnace (three electrodes with alternating current). The electrodes should be of ample cross section to carry the current without overheating because of resistance. 7 Reference has already been made to the neces l. sity for using finely divided copper in the prelim-- inary reduction. That the factor of metal coarseness orflneness plays an important role in this matter, fromthe standpoint of .optimum results,

can'be judged from the following quantitative ll testsyall being run under conditions as identical as possible other than for the'changes in the particle size of the copper used. I I

A beryllium oxide containing 3.04% AlzOa, 0.22% s1o=, 0.2% F8203 was used throughout. 8 Witha relatively coarse copper powder, varying in mesh from 38 to 60 to the inch, an alloy of only 1.47% beryllium was obtained, with aluminum at 0.20%. With copper at 60 to 100 mesh in particle size an alloy of 2.68% beryllium was obtained, aluminum being 0.26%. when identical reduction was accomplished, but using a copper powder of still finer size, specifically 100 to 200 mesh,.beryllium was 2.72%, and aluminum remained the same as before. When copper powder ran to finer than 200 mesh, approaching a copper dust, the beryllium content of the resulting alloy dropped somewhat, to 2.38%, while aluminum rose to 0.34%. In all cases reported, 2 silicon was 0.l1|8%-0.10% and iron from 0.07%- 0.09% indicating that particle size does not signiflcantly affect their ratio in the final alloy.

It will be seen that there is a deflate-preferability for copper powder of smaller than 60 mesh 0 size. Beyond this, extreme fineness does not significantly increase the yield. Because the percentage of impurities remains constant in the final alloy, it follows that the ratio of impurities to beryllium is relatively higher in alloys corresponding to the use of coarse powder.

In all, the optimum alloy is to be obtained by use of a copper powder of less than 60 mesh average particle size. Where, therefore, the term finely divided copper, or its equivalent, is used in this specification or in the claims, it is to be understood as referring tosuchmetal preferably.

less than 60 mesh average in particle size.

The invention is not to be limited to copperberyllium alloys, but is intended to include other 55 oxide, carbon and copper, the mass providing'suf- 05 I ficient reaction material laterally around the arc .to produce a drop in temperature to at least about 1285 C., and sufllcient. reaction'material above the arc to trapberyllium vapor, maintain--v f tainingthe temperature below the are below the 701 pointwhere the beryllium vapor pressure causes material loss, and adding massive copper to the reaction'mass in an upper region thereof after preliminary reduction-to 2 I1 1 theprocess of .forminKberyHium-copper,

ash down alloy formed.

alloys by thermal carbon reduction of beryllium oxide in the electric arc furnace, the step of adding massive copper to the reaction mass subsequent to preliminary'reduction in the presence of j copper of less than 60 mesh average particle size.

3. The process of continuous formation of beryllium-copper alloys characterized by the ANDREW J. GAHAGAN.