459

Kupfer

· 59. Jahrgang · 11/2005

standing corrosionperformance are tak-en advantage off.As starting materialfor Copper productsCopper cathodeshave the highestpurity. This materialcan only be degradedduring remelting byabsorbing gases andsolid inclusions. Insecondary materialsbesides dissolvedgases also alloyingelements can be pre-sent as dissolvedimpurities that lowerthe properties. In thispaper melt treatmenttechnologies for therefining of secondaryCopper melts are dis-cussed. Especiallythe removal of gases

from secondary Copper, the removalof dissolved metallic impurities aswell as the treatment of solid inclu-sions are reviewed in detail.

Copper high tech applications

As mentioned above, Copper is main-ly used for electrical conductivity

Besides the use as electricalconductor, Copper is used in avariety of other applica-tions.

Major fields are fresh water tubing,roofing, heat exchangers, as mouldsfor continuous casting of Aluminiumand Copper, in power generators, inwire etc. In these cases other positivecharacteristics of Copper like superi-or heat conductivity and an out-

applications. Besides the day-to-dayhousehold uses there are some fieldswhere extreme product cleanliness isnecessary. One example is the use asa cladding material for superconduc-tors as shown in Figure 1. In case ofcooling failure (most superconductorsstill need very low temperatures) theCopper matrix takes over the conduc-tion because of good deep tempera-ture conductivity until the system isshut off. Contrarily to the supercon-ductors Copper does not reveal asharp resistance step at a certain tem-perature.Another example for Copper in high-tech industries is the application inthe micro electronics industry. Theelectronic industry boomed with thedevelopment of the so called printedcircuit boards (PCBs): a Copper foil isapplied onto a non conducting sub-strate; alternatively the Copper isdeposited electrolytically on the sub-strate. Then the circuit lines are print-ed with a special colour, the excessCopper is etched away and the con-ductor lines remain. Nowadays layersof only 50 µm Copper are applied.The boards with these thin Copperlayers can also be made flexible. Thisis a great challenge for the material:it has to combine a high conductivitywith good mechanical properties andelasticity.One last example for a Copper high-tech application is the use in connec-tor pins, an every day’s, but often notrecognised application. In modernautomobiles numberless electronichelpers are connected to one centralsite. Therefore tens of connectors arecombined in one connector system.The pins need a high conductivity butalso a flexibility that allows toassemble and disassemble them often.At the same time they should be stiffenough to guarantee the electricalcontact even at temperatures in an

Melt treatment of Copper -„Ways to a high tech material“

Friedrich, B.; Kräutlein, C.; Krone, K. (1)

Copper is presumably the first metal used tech-nically by mankind. First archaeological disco-veries date back 10.000 years. During time theapplication fields for Copper and its alloysdiversified and changed drastically, especially inthe last century. The biggest change of its usewas initiated with the discovery of electricity;Copper is the second best electric conductorbesides Silver. Today the major part of Copper(approx. 50%) is used in conductivity applicati-ons, either as pure metal in wires or as alloys inconnectors. Early in the development phase itwas discovered, that even very small amounts ofimpurities decrease its electrical conductivitysignificantly. In 1913 the International AnnealedCopper Standard (IACS) was defined with 58MS/m as minimum conductivity for high con-ductivity Copper [1]. Today qualities with102.5% IACS and more are available. An in-crease of conductivity leads to a decrease inohmic heat generation, this is especially impor-tant in the microelectronics industry.

Figure 1: Copper in super conduc-tors applications

· 59. Jahrgang · 11/2005460

Metal l-Forschung

engine compartment and should havethe same lifetime expectance as theautomobile.These three fields of application are,of course, only a very small extract ofthe Copper world. Copper and itsalloys are also used in vacuumswitches, vacuum capacitors, electronbeam tubes, welding electrodes, heatexchangers, as moulds for con-tinu-ous casting of steel, Aluminium andCopper, in generators of power plantsand wire and so on.

Processing of liquid Copper

The processing of primary Copperusually starts with the melting ofCopper cathodes in a shaft furnace.For melting of Copper scrap in gener-al induction furnaces or drum fur-naces are applied. After the prelimi-nary melting furnace the melt iseither casted directly or subjected to acasting furnace where the melt isstored, (alloyed) and heated to cast-ing temperature. For Copper the con-tinuous wire casting by castingwheels or Hazelett casters is especial-ly important. Besides wire castingalso vertical and horizontal slab cast-ing is applied as well as mould cast-ing etc.

Impurities in Copper melts

The issue of impurities in Copper canbe separated in two parts: impuritiesin primary Copper remaining or col-lected after refining electrolysis andimpurities in secondary not electro-refined Copper scrap. The refiningelectrolysis produces cathodes withmin. 99.995 wt. % Cu, the majorremaining impurities are Silver, Sul-phur, Nickel and Iron. But the con-tents are usually so small that theyare not detrimental to the propertiesof Copper. The more critical elementsin this sense namely Hydrogen andOxygen as well as inclusions enterthe primary Copper usually duringthe remelting and casting process.In secondary materials the impuritymatter is more complicated. Remelt-ing of Copper scrap makes sense eco-logically and economically becausethe material does not have to be leadback into the energy intensive pri-mary electrolysis. There are twotypes of scrap, the sorted and mostlyclean production scrap which is eas-ily and directly reusable and end oflifecycle scrap (=”old scrap”) consist-ing often of a mixture of differentalloys or even compounds with othermetals and materials. In the processof producing a clean and specifiedalloy the undesired elements eitherhave to be removed or diluted. Incase they are not removed they cane.g. form intermetallic phases in theCopper matrix and as a result

decrease the mechanical propertieslike the ultimate yield strength andthe ductility. Due to the noble char-acter of Copper, most elements likeSilicon, Aluminium and Iron caneasily be removed from a Coppermelt by selective oxidation up to avery low concentration (activity).Physically and chemically more sim-ilar elements like Nickel, Cobalt, Tinand Lead have to be treated withmore attention. Elements and com-pounds typically occurring in Coppermelts are summarised in Fig.ure 4.Dissolved metallic impurities in smallamounts mostly are not detrimentalto the properties of Copper, but someelements as for example Lead andArsenic precipitate at the grainboundaries of the Copper materialsand lead to embrittlement of thematerial.Generally an Oxygen and Hydrogenpick-up can lead to very negativeeffects. The two gases have a highsolubility in liquid Copper thatdecreases sharply during solidifica-tion. This can lead to a bubble for-mation, i.e. porosity in the solidmaterial. Oxygen can also formcuprous oxide (Cu2O) above its solu-bility level that immediately reactswith the hydrogen of an annealing orwelding atmosphere forming watervapour during annealing or welding,this phenomenon is called Hydrogenembrittlement. Dissolved Hydrogenand Oxygen (or Cu2O) will react towater under extreme pressure in the

Figure 2: Copper strip andconnectors

Figure 3: Processing of molten Copper to semi finished products

461

Kupfer

· 59. Jahrgang · 11/2005

lattice and will form cracks and leadto embrittlement.Solid inclusions like intermetallics oroxides from alloying elements or therefractory material usually do nothave a negative impact on Copperand Copper alloys. Because the den-sity difference between the Coppermelt and the particles is very highthe particles tend to float to the sur-face (e.g. the density of Copper at1100°C is 7.96 g/cm3 while Iron oxidehas a density of 5.25 g/cm3 [2]).However Stokes law predicts thateven at high density differences verysmall particles tend to stay suspend-ed (In case of FeO particles the diam-eter smaller than 10 µm rise with0.59 m/h). This leads to problemswhere wires with a very small diam-eter are drawn, for thin Copper foilsor where small connector pins aremade from thin Copper strip, beingetched or punched.

Requirements for cleanCopper melts

The most frequent impurity in refinedCopper is Oxygen. If the alloy is notmolten un-der Oxygen free atmos-phere, Copper melts will always pickup Oxygen. Copper melts hyper satu-rated with Oxygen lead holes or theso called Hydrogen embrittlementduring casting and welding in caseHydrogen is present in the material orthe atmosphere. Dissolved Oxygencan also oxidise less noble elementswhich leads to the formation of brit-tle oxides in the material. In techni-cally alloys, an Oxygen content of afew hundred ppm is accepted where-as in special grade materials the con-tent has to be lowered below oneppm, some grades and their respec-

tive tolerated Oxygen content areshown in Figure 5.In technical alloys the limit for spec-ified metallic impurities mainly variesbetween 0.03 wt. % and 0.05 wt. %.A typical requirement for DHP-Cu isfor example all ele-ments except forPhosphorous <0.05 wt. %, specifiedelements (Fe, Ni, Zn) max. 150 ppmall remaining elements max 50 ppm[3]. For high quality products lim-itsof < 90 ppm Pb, < 50 ppm Sn and <150 ppm Ni are required.Solid inclusions are especially aproblem in thin wire drawing andthin sheets rolling. Equation 1describes the critical inclusion size,any particle bigger will initiate wirebreaks:

Eq. 1with dcritical: inclusion size that leadsto a break, D: wire diameter, K is amaterial con-stant, T: applied pullingtension, n: reduction ratio.That means that the possibility for awire break increases sharply with thesize reduction ratio and the pullingtension. In foil production inclusions

lead to a foil break in the worst case.Smaller inclusions lead to surfacedefects and elongated holes in thematerial.

Refining by melt treatment

The mentioned impurities can bereduced to an uncritical content byan appropriate melt treatment beforecasting. Different melt treatmenttechnologies of Copper were devel-oped for different impurities. The“classical” melt treatment of Copperis the oxidation by air or Oxygeninjection or top blowing. With thistechnique elements that are lessnoble than Copper can be removedfrom the melt. Today this technique isoften combined with a specific slagthat can take up the impurity oxidesand supplies a certain Oxygen poten-tial to improve the impurity separa-tion.To remove dissolved gases especiallyHydrogen and Oxygen the oldestmethod is first of all the right melthandling. That means that the openmelt surface has to be shielded andmelt transfer has to be in a laminarflow in order to inhibit turbulences.Hydrogen can be removed by anexcess of Oxygen in the melt. Afterthe removal of Hydrogen by oxidis-ing conditions, the melt has to betreated under reducing conditions toremove the Oxygen. A new pick-upof Hydrogen has to be avoided byshielding of the melt with slags, coallayers or shielding gases. Oxygen isusually added by blowing air on themelt surface. For the reduction since

Figure 4: Impurities in Copper melts

Figure 5: Requirements for Oxygen contents in Copper

former times the so called poling bytree trunks, especially birch, is used[4]. This technique is still used todayin some places. An advantage of thetree trunks is that they have a CO2emission of zero, as plants are con-sidered to be regenerative. Alterna-tively to this classical procedure,modern techniques reduce the partialpressure in the surrounding atmos-phere; this led to vacuum and gaspurging technologies. A flow chart ofpossible melt treatment routs isshown in Figure 6.The removal of dissolved metals inCopper is an upcoming problembecause of the increasing recyclingmaterial volume not being treated byrefining electrolysis. Metals like Zinc,Arsenic and Antimony can be evapo-rated by a vacuum treatment, for oth-ers like Nickel, Cobalt, Tin and Lead aspecial slag treatment is more eco-nomic. Nevertheless very often thehigh specifications of high techapplications cannot be met usingonly one refining technique.To remove solid particles from Cop-per melts a simple settling is usuallyenough for standard qualities. But theincreasing requirements, e.g. for wireproduction, led to the development offiltration and floatation techniques.Typical filter materials made of alu-mina or silicon carbide are shown inFigure 7.

All the melt treatment techniques areactually batch processes. They haveto be im-plemented in the existingprocess routes in a way that theireffect is not lost before casting andsolidification of the metal. Thatmeans that after deoxidation, a pickup of Oxygen has to be avoided byproper casting gutters protected by acoal or coke cover or a shielding gas.After filtration a laminar flow of themelt through the launder has to beassured to avoid turbulences thatpromote abrasion of the refractoryand formation of oxides.

Deoxidation

A deoxidation has to be conductedafter all melt treatments that enclosean oxidative refining. As deoxidationagents be applied gases, certainreagents and electricity.

Deoxidation by gas purging

Gas purging of Copper alloys can bedivided into purging with inert gasesand purg-ing with reactive gases. Thepurging with inert gases is based on alow partial pres-sure of the gas thatneeds to be removed. This process isdiffusion controlled, i.e. the speeddepends on the diffusion constantand the specific surface area of themelt-bubble interface. A diffusioncontrolled mass transfer can be influ-

· 59. Jahrgang · 11/2005

enced by rising temperatures (techni-cally not feasible) and a decrease inthe thickness of the Nernst layer.Thus the process can be significantlyintensified by an increase of the sur-face of the gas bubbles by an appro-priate gas supplying technique. Theprinciples for the gas purging with aninert gas was described by Friedrichet. al. [5] elsewhere. Argon and Nitro-gen are appropriate inert gases forpurging Copper, especially for Hydro-gen removal. Their solubility in liquidCopper is negligible.Gas purging methods for Copperdeoxidation were developed toreplace the poling with tree trunksthat goes back at least to the year1200 AD when it was mentioned in“De Re Metallica”. In the 1960s exten-sive research was conducted on reac-tive gases with different gaseous andsolid reducing agents. Reducing gas-es were tested like different carbon-hydrates (Methane, Propane, Butane)[6], Carbon monoxide as well asHydrogen. Ammonia for gas purgingwas investigated in the 1970s [7].Also oil or coal dust were used asreducing agents, the problem of thesesubstances was their impurity con-tent, especially Sulphur that enrichesin the Copper melt.Natural gas as a reducing agent wastested by Klein [8] but the processwas found to be very ineffective andslow. In comparison a reformed, par-tially oxidised natural gas where thecarbon hydrates are reacted to carbonmonoxide and Hydrogen leads to afast deoxidation of the Copper melt.This process was implemented at the

462

Metal l-Forschung

Figure 6: Flow chart of possible melt treatment steps

Figure 7: Typical CFF filter platesfor melt filtration (CFF= ceramicfoam filter)

Phelps Dodge refining works in the1960s according to US-patent2,989,397. An intensive investigationon the kinetics of the Copper deoxi-dation by carbon monoxide was car-ried out by Andreini et al. in 1977 [9].The chemical gross equation of thedeoxidation with carbon monoxide is

Eq. 2It was found that the Oxygen diffu-sion in the liquid Copper to the gasbubbles is the rate controlling step,valid in a concentration interval of50 to 1000 ppm. Below 50 ppm thekinetics of the Oxygen removaldecreases sharply. This can not be ex-plained by equilibrium reasons, it ispossibly due to interactions with Sul-phur in the melt. This study agreesfairly well with an older study fromNanda et al [10] which also found asharp decrease of the deoxidationspeed below a concentration of 50ppm Oxygen. The lowest Oxygenconcentration in liquid Copper thatcould be achieved by this process was10 ppm.As an alternative to carbon hydrates,ammonia was investigated as a possi-ble re-ducing agent by Henych etal.[7]. The gross equation of this reac-tion is:

Eq. 3The reducing element in this case isthe Hydrogen. The developing Nitro-gen is an inert gas and does not dis-solve in the Copper melt. The lowestOxygen concentration in liquid Cop-per that could be achieved by thisprocess was 200 ppm. Ammonia isvery expensive and at room tempera-ture a liquid, therefore it has to begasified which increases the com-plexity of the process. Therefore thistechnology never achieved industrialscale.It is also possible to use a Nitrogen-Hydrogen mixture instead of ammo-nia. The ad-vantage is that the purg-ing media are already gaseous [11].But pure Hydrogen is quite expensiveand mainly produced from naturalgas, so there is no advantage in com-parison to natural gas.Up till now the favoured gas injectiontechnology in the Copper industry is

tuyeres/injectors even though in oth-er industries different gas supplyingtechniques are used as rotary vaneddispersers and porous plugs.

Deoxidation by addition of solidadditives

Besides deoxidation with reactivegases also deoxidation procedureswith a solid element- or compound-additive exist. For Copper the inser-tion of Phosphorous, Lith-ium, Calci-um, Boron, Beryllium, Aluminium,Silicon, Magnesium and Zinc as wellas Calcium-Boride [11];[13] wereinvestigated starting in the 1930s.The deoxidising agents are added instoichiometric amounts or with avery slight excess related to Oxygen.The oxides of the added elementsshould evaporate, float and form ordissolve in a slag. Any excess agentsunfortunately dissolve in Copper orform compounds and harm the prop-erties of the material. For examplePhosphorus leads to a decrease in theconductivity of Copper as shown inTable 1. Lithium in Copper does notresult in such a drastic loss of con-ductivity but it is rather expensiveand reacts with refractories and car-bon bearing shielding layers. Somemultiple-step processes were devel-

oped for exotic applications, e.g. twostep processes with Zinc and Magne-sium, Zinc and Beryllium, Calciumboride and Magnesium, Phosphorousand Lithium or Calcium boride aswell as processes with three steps asfor example Zinc, Calcium boride andBeryllium [13]. With these techniquesOxygen contents of below one ppmcan be obtained. But the lower theachieved Oxygen content is, thehigher is the concentration of the dis-solved deoxidation agent causing alower conductivity. Table 1 Showsrelevant properties of possible deoxi-dation agents.The general equation of the deoxida-tion of Copper melts by solid reduc-ing reactants (R) is

Eq. 4Most of the oxidic compounds result-ing from deoxidation have a highmelting point. So as a result besidesthe deoxidation treatment also a slagforming constituent like SiO2/FeO/CaO has to be added to obtain a slagwith a low viscosity and melting tem-perature that can be easily and entire-ly removed. The only elements form-ing low melting oxides are Boron andPhosphorus. Because of the higherprice of Boron in comparison to Phos-

463

Kupfer

· 59. Jahrgang · 11/2005

Table 1: Properties of deoxidising agents and their oxides[2];[13];[14]

Element Resultingresistance at0.1 wt. % in

Cu

Max.solubilityin Copper

Oxidecompoundformed by

deoxidation

Density ofthe oxide

Meltingtemperatureof the oxide

/µΩcm /wt. % /g/cm3 /°C

Al 1.89 5.7 Al2O3 3.423 2290

B 2.435 0.3 B2O3 2.44 (hex) 460

Be 2.1 16.4 BeO 3.02 2575

Ca 1.726 <0.001 CaO 3.4 2927

Li 2.2 2.8 Li2O 2.013 1560

Mg 1.876 3 MgO 3.576 2831

P 3.04 2 P2O5 2.93 580

Si 2.344 5 SiO2 2.648 1550

Zn 1.709 38.27 ZnO 5.66 1975

Cu 1.5 - - 8.96 1085

O 1.673 at0.01 wt%

20*10-4 Cu2O 6.00 1244

phorous, Phosphorous is the deoxida-tion element of choice until today.Phosphorous is usually added as amaster alloy with the composition90% Cu and 10% P in the form ofwaffle plates, which are dumped intothe furnace. For a fine adjustment ofthe P-content during continuous cast-ing Phosphorous may be added incored wire or as granulates.In practice for deoxidation multiplesteps are applied as for exampledeoxidation by tree trunks, followedby deoxidation with phosphorousand then shielding by carbon layerson top of the melt and burners withreducing conditions.

Electrochemical deoxidation

Besides the known and proven deox-idation techniques more recently gal-vanic deoxidation was investigated.With Calcium oxide doped Zirconiumoxide was proposed as a high temper-ature oxygen ion conductor alreadyin 1957 [15]. The technique for elec-trochemical oxidation or reductionwas firstly used to determine the oxy-gen solubility in metal melts espe-cially in Copper [16].The electrochemical that is appliedcell can be written as Cu ZrO2 + CaO(Y2O3) Pt, airO (at p´O2) solid electrolyte(p´´O2=0,21 atm)The heart of this technique is the sol-id electrolyte oxygen ion conductor.Through this conductor an Oxygenexchange between a closed melt andfor example the atmosphere can hap-pen but only in the ionic form. So atone side of the electrolyte the Oxygenis supplied with electrons and isionised to O2- and at the other side itleaves electrons. For charge equilibri-um the electrons have to move theother way. This can only happen byan electron conductor. The electronmovement results in a voltage thatcan be measured, it is dependent onthe oxygen concentration differencebetween melt and surroundingatmosphere. In case the voltage iszero there is no Oxygen concentra-tion difference between melt andatmosphere. Vice versa if a voltage issupplied externally oxygen can bepumped in or out of a melt. The

scheme is shown in Figure 8. In casethe oxygen exchange is due to con-centration differences, the process iscalled galvanic oxidation/reduction,in case a voltage is supplied it is anelectrochemical oxidation/reduction.On this principle are based also forexample the Lambda probe in cars orthe technology of fuel cells. It is alsovery common for batch wise oxygenmeasurements in Copper melts.In the 1970´s first trials for deoxida-tion of metal melts were conducted.In Copper melts a value of 7 ppm OCu

could be achieved. It was soon dis-covered that diffusion is not fastenough to supply the electrolyte withsufficient Oxygen and sever me-chanical or inductive stirring is nec-essary to achieve a fast deoxidation.R. R. Odle achieved on laboratoryscale a reduction from 500 ppm OCu

to 25 ppm OCu in 80 min [16].In the end of the 1990´s pilot scaletest runs for electrochemical deoxida-tion cells were conducted by P. Soral[16]. In this pilot scale unit as elec-trolyte Yttria stabilised Zirconia wasused (YSZ) with a Ni-YSZ cermet asan anodic electron conductor on theinside of electrolyte and stainlesssteel rods as the cathodic current sup-plier in the melt. It was found that theelectrolyte tubes are very prone tothermal shock and therefore a notcomplete deoxidation was obtained.But future research is ongoing to findmore thermal shock resistant elec-trolytes for the use in industrial envi-ronments.

Impurity removal by slag treat-ment

Slags in Copper refining have thetask to take up the impurity oxides,that means they should have the fol-lowing properties:

High solubility for impurityoxidesLow solubility for Copper andCopper oxideLow melting temperature closeto the melting temperature ofCopperHigh thermal stabilityLow interaction with the refrac-tory material

· 59. Jahrgang · 11/2005

Before or during a slag treatment theCopper melt has to be oxidised forexample by top blowing of air. A slagthat is industrially used is the socalled “Fayalite” slag that is alsoapplied in the primary flash smeltingfurnace. It is based on the system FeO– Fe2O3 – SiO2. It is effective for theremoval of Cd, Fe, Pb, Sn and Znespecially for elements with thevalence of two that can be trapped assilicates. This slag can be used in fur-naces with silica refractory [16]. Thisis the slag mainly used today.On laboratory scale other slag typeswere investigated as for example Cal-cium-ferrite slags that are based onthe ternary system FeO – Fe2O3 –CaO. They remove the elements Al,As, Fe, Sb and Sn, especially elementswhich exhibit an acidic character in aslag [16] at their highest oxidationlevels. Another system is based onCaF2 – CaO – MgO – SiO2 this slagshows the same behaviour as the“Calcium-ferrite” slag but with a low-er solubility for Copper oxide.Oxide slags often contain by-prod-ucts from the deoxidation treatmentwith solid reactants. That meansespecially B2O3 and P2O5. These com-pounds can further influence theactivity of impurity metals like Lead,Nickel and Tin. Here the research is-ongoing on the influence of thesecompounds.As slag additions for oxide slagsbesides pure oxides like SiO2 or CaOalso so called residues can be appliedas for example red mud from thebauxite processing during Alumini-um production.Also investigated by researchers onlaboratory scale were salt slags that

464

Metal l-Forschung

Figure 8: Scheme of a electroche-mical cell with solid electrolyte

are mainly based on Sodium carbon-ate (Na2CO3) but also on other alka-line carbonates like lithium andPotassium carbonates (Li2CO3) and(K2CO3). They are very effective refin-ing slags, but they all attack “usual”refractories of the Copper industry. Itwas found that by fluxing with asodium carbonate slag the amount ofarsenic and antimony could be low-ered below 0.1 ppm. The problem thatoccurred was, that the binary solutionof Sodium carbonate and Antimonyoxide let to an increase of solubilityof Copper oxide in the slag and there-fore let to higher Copper losses [21]. For fluxes on other salt bases mainlyfluorides can be used because theirmelting and evaporation point is veryhigh. Here especially systems on Cal-cium fluoride (CaF2) base with a mix-ture of Aluminium fluoride (AlF3) orSodium fluoride (NaF) are possiblecandidates. All salt fluxes operate inthe best way when the reaction sur-face between metal and slag isincreased for example by solid fluxinjection in the metal melt [23]instead of an addition to the surface.

Vacuum treatment

The vacuum technology in the Cop-per industry is mostly used for theproduction of low alloyed alloys withreactive elements and for the produc-tion of master alloys on Copper basewith reactive elements like Chromi-um, Zirconium, Magnesium and Be-ryllium. Qualities with an Oxygencontent below one ppm are also pro-duced by this method [27]. Theamount of vacuum produced Coppermaterials is very small com-pared tothe total amount of produced Copper.Vacuum treatment for special prod-ucts of Copper alloys mainly is usedto remove dissolved gases like Hydro-gen and Oxygen or less frequently toevaporate volatile elements like Zinc,Bismuth and Arsenic etc. First trialsfor the removal of Oxygen, Hydrogenand Sulphur from Copper melts wereconducted by R.A. Stauffer et. al. in1948 [24]. They started by meltingCopper in a retort using a graphiteboat. The carbon is necessary becausethe decrease of the pressure during

the vacuum treatment is not suffi-cient to remove Oxygen to low val-ues, a CO formation/removal is muchmore effective. After three hours thechemical analysis showed a reductionof Hydrogen to 0.1 ppm, of Oxygento 3.8 ppm and of Sulphur to 1.9ppm. Despite these good results thevacuum technology for gas removalwas never implemented for the pro-duction of technical materials due thehigh investment costs for vacuumequipment and its maintenance. Onlythe company Hitachi patented a con-tinuous process in 1989 that is in useuntil today [21].In the 1970s the possibility to removeother elements than dissolved gasesby means of vacuum was investigat-ed. In the focus were mainly Bismuth,Antimony and Arsenic because theyoften accompany the Copper ores andare difficult to remove during therefining electrolysis. These elementshave a negative effect on the proper-ties of Copper and its alloys, so theirremoval is of great importance [26].The removal of Bismuth, Lead,Arsenic and Antimony from Copper[26] was investigated and it wasfound that for Copper alloys the Bis-muth content could be reduced to 70ppm, Arsenic to 1260 ppm and Anti-mony to 310 ppm. The distillation ofvolatile elements is not very commonindustrially.For the removal of gases three differ-ent setups of furnaces can be used.

Usually the heating source in vacuumfurnaces is an induction heating. Thevacuum induction melting (VIM) fur-nace consists of a cold walled cham-ber in which all equipment likemoulds etc. is stored. The inductioncoil can be tilted so that inside thematerial can be casted inside the fur-nace under vacuum. In the vacuuminduction degassing system (VID) alid is positioned on the top of the fur-nace. Before casting the lid has to beremoved, and the melt is subjectedagain to the atmosphere. The advan-tage of this system is the small size incomparison to the VIM. The third sys-tem is the so called vacuum inductiondegassing pouring (VIDP). The coldwalled container of the induction coilcan be tilted so that the melt flows ina launder and leaves the cold walledsystem through a closed launderchamber. This system avoids the con-tact to the air during casting andcombines the advantages of VIM andVID.

Summary

In this paper the possibilities for themelt treatment of Copper and Copperalloy melts are discussed. Gas purg-ing, deoxidation by chemical reac-tion, slag treatment and vacuumtreatment were reviewed in moredetail. The review shows that woodpoling can achieve oxygen contentsto a few hundred ppm, cracked natur-al gas can reduce the Oxygen downto a few tens of a ppm and a deoxi-dation by solid reactants leads tobelow 1 ppm of Oxygen in a Coppermelt. A vacuum treatment can alsoreduce the oxygen level to ~ 1 ppm.With a melt treatment with a slag dis-solved metallic impurities can bereduced below critical values. Fay-alite slags that are the most commoncan remove elements with thevalence of two as silicates as forexample Pb, Cd, Sn and Fe. Carbon-ate slags can remove As and Sbbelow 1 ppm, but are not technicallyused.With The results show that nearly allelements can be removed from Cop-per melts in the molten state, avoid-ing the energy intensive refining

465

Kupfer

· 59. Jahrgang · 11/2005

Figure 9: Vacuum furnace [27](VIM)

electrolysis. Because all melt treat-ment technologies in the Copperindustry are batch processes theirimplementation has to undergo criti-cal consideration to fit them into theexisting process lines. Also the effectof the different treatment technolo-gies has to be maintained until themetal is casted and chilled. Thereforeeffective shielding measures by car-bon layers and shielding gases haveto be added to the actual melt treat-ment process. The state of the art of the existingmelt treatment process technologiesenables the Copper industry toachieve the dire demands of the semifinished product consumers.

References[1] G. Armstrong Smith: “Some Observa-

tions on the electrical conductivity ofcommercial electrorefined Copper“;Journal of the Institute of Metals, Vol.100, 1972, pp. 125 – 130

[2] R. Blachnik, D´Ans Lax: “Tachenbuchfür Chemiker und Physiker: Elemente,anorganische Verbindungen und Materi-alien, Minerale“; Springer Verlag 1998,ISBN 3-540-60035-3

[3] CEN/TS 13388:2004[4] E. Brunhuber: “Schmelz- und Legierung-

stechnik von Kupferwerkstoffen”; 1959,Fachverlag Schiele und Schön GmbH

[5] B. Friedrich, K. Krone, C. Kräutlein:“Melt-Treatment of Aluminium – Waysto a high performance Metal“; Metall,Vol. 59, 2005, pp. 30 – 36

[6] F.E. Brantley, C.H. Schack: Bureau ofmines R.I. 6113, 1962

[7] R. Henych, F. Kadlec, V. Sedlacek: “Cop-per refining by gaseous Ammonia“; Jour-nal of Metals, 1965, pp. 386 – 388

[8] L. Klein: “Gaseous reduction of Oxygen-containing Copper”; Journal of Met-als,August 1961, pp. 545 – 547

[9] R. J. Andreini, J.S. Foster, R. B. Phillips:“Copper deoxidation kinetics utilizingcarbon monoxide“; MetallurgicalTransactions B, Vol. 8B, 1977, pp. 633 –638

[10]C. R. Nanda, G. H. Geiger: “The kinet-ics of deoxidation of Copper and Cop-per alloys by carbon monoxide“; Metal-lurgical Transactions, Vol. 2, April 1971,pp. 1101 – 1106

[11] K. Fasshauer, F. Steffner, H.-J. Dauerst-edt; “Verfahren zum Polen von Kupfermit Wasserstoff und Stickstoff”; Euro-pean Patent 0992597B1, 2002

[12]Kh. G. schmitt-Thomas, H. Meisel, H.-J. Dorn, H. Rasavisadeh: “Einfluss derKalziumhexaborid (CaB6)-Desoxida-tion auf elektrische Kennwerte vonKupfer“; Metall, Jhrg. 29, Heft 12,Dezember 1975, pp. 1198 – 1204

[13]E. Brunhuber: “Die Herstellung elek-trisch hochleitfähiger Gussstücke ausKupfer und Kupferlegierungen”;Giesserei, Januar 1959, pp. 2 – 8

[14] Predel, B.; in: Landolt-Börnstein -Group IV; Vol. 5x, xxxx, Springer-Ver-lag Heidelberg, ISSN 1616-9557

[15]K. Kiukolla, C. Wagner; “Galvanic Cellsfor the determination of the standard mo-lar free energy of formation of metalHalides, Oxides, and Sulfides at elevat-ed temperatures”; Journal of the electro-chemical society, 1957, pp. 308 - 316

[16]P. Soral, U. Pal, H. R: Larson, B.Schroeder; “A pilot-scale trial of an im-proved galvanic deoxidation process forrefining molten copper”; Metallurgicaland Materials Transactions B, 1999, pp.307 - 321

[17]R. R: Odle, R. A. Rapp; “Further studyof the electrochemical deoxidation of in-

· 59. Jahrgang · 11/2005

duction-stirred Copper melts”; Metallur-gical Transactions B, 1977, pp. 581 -589

[18]D. Janke; “Elektrolytische Desoxyda-tion von Kobald-, Nickel-, Kupfer- undSilberschmelzen”; Zeitschrift für Metal-lkunde, 1977, pp. 302 - 307

[19]J. Gerlach, J. Osterwald, W. Stichel;“Coulometrische Bestimmung der Mis-chungslücke zwischen flüssigem Kupferund Kupfer(I)-oxid”; Zeitschrift fürMetallkunde, 1968, 576 - 582

[20]J. Gortais, F. Hodaj, M. Allibert, J.-M.Welter: “Equilibrium Distribution of Fe,Ni, Sb and Sn“; Metallurgical and Mate-rials Transactions B, Vol. 25B, Oktober1994, pp. 645 – 651

[21]Japanese Patent JP01142016A, 1989[22]H. Fukuyama, T. Fujisawa, C. Ya-

mauchi: “Removal of Antimony and Ar-senic from molten Copper by SodiumCarbonate Slag Treatment“; Metallurgi-cal Processes for the Early First Centu-ry, The Minerals, Metals & Materials So-ci-ety 1994, pp. 443 – 452

[23]B. Zhao, N. J. Themelis: “Kinetics of Asand Sb removal from molten Cop-per“;EPD Congress, The Minerals, Metals &Materials Society, 1995, pp. 515 – 524

[24]R. A. Stauffer, K. Fox, W. O. DiPietro:“Vacuum melting and casting of Cop-per“; Industrial and Engineering Chem-istry, Mai 1948, pp. 820 – 825

[25]O. Yutaka, O. Takeji, Japanese Patent01142016, 1989

[26]R. Harris: “Vacuum refining Coppermelts to remove Bismuth, Arsenic andAntimony“; Metallurgical TransactionsB; Vol. 15B, Juni 1984, pp. 251 – 257

[27]H. Rupp: “Anwendung der Vakuum- undSondermetallurgie für Kupfer und seineLegierungen“; Metall. Vol. 42, Heft 4,April 1988, pp. 356 - 363

(1) Prof. Dr.-Ing B. Friedrich, Dipl.-Ing. Christoph Kräutlein, Dr.-Ing.K. Krone †, RWTH Aachen

466

Metal l-Forschung