c
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 10.1002/14356007.a07 471
Copper
Adalbert Lossin, Norddeutsche Affinerie Aktiengesellschaft, Hamburg, Federal Republic of Germany
1. Introduction . . . . 2
2. Physical Properties . . . . 4
3. Chemical Properties . . . . 6
4. Occurrence . . . . 8
4.1. Copper Minerals . . . . 8
4.2. Origin of Copper Ores . . . . 9
4.3. Copper Ore Deposits . . . . 10
4.4. Copper Resources . . . . 10 4.5. Mining . . . . 10 5. Production . . . . 11 5.1. Beneficiation . . . . 12 5.2. Roasting . . . . 15 5.3. Pyrometallurgical Principles . . . . . 17
5.3.1. Behavior of the Components . . . . 17
5.3.2. Matte . . . . 17
5.3.3. Slags . . . . 17
5.3.4. Oxidizing Smelting Processes . . . . . 19
5.3.5. Proposals . . . . 19
5.4. Traditional Bath Smelting . . . . 21
5.4.1. Blast Furnace Smelting . . . . 21
5.4.2. Reverberatory Furnace Smelting . . . 22
5.4.3. Electric Furnace Smelting . . . . 22
5.4.4. Isasmelt Furnace . . . . 24 5.4.5. Noranda Process . . . . 24 5.4.6. CMT/Teniente Process . . . . 25 5.4.7. Vanyukov Process . . . . 25 5.4.8. Baiyin Process . . . . 26 5.5. Autogenous Smelting . . . . 26
5.5.1. Outokumpu Flash Smelting . . . . 27
5.5.2. Inco Flash Smelting . . . . 29
5.5.3. KIVCET Cyclone Smelting . . . . 30
5.5.4. Contop Matte Smelting . . . . 30
5.5.5. Flame Cyclone Smelting . . . . 31
5.6. Discontinuous Matte Conversion . . 32
5.7. Continuous Matte Conversion . . . . 35
5.7.1. Noranda Process . . . . 35
5.7.2. Mitsubishi Process . . . . 36
5.7.3. Kennecott/Outokumpu Flash Convert-ing Process . . . . 38
5.8. Direct Blister Smelting . . . . 38
5.8.1. Blister Flash Smelting . . . . 38
5.8.2. QS Process . . . . 39
5.9. Copper Recycling . . . . 39
5.10. Hydrometallurgical Extraction . . . 40
6. Refining . . . . 45
6.1. Pyrometallurgical Refining . . . . 45
6.1.1. Discontinuous Fire Refining . . . . 45
6.1.2. Continuous Fire Refining . . . . 46
6.1.3. Casting of Anodes . . . . 46
6.2. Electrolytic Refining . . . . 47
6.2.1. Principles . . . . 47
6.2.2. Practice of Electrorefining . . . . 49
6.3. Melting and Casting . . . . 50
6.3.1. Remelting of Cathodes . . . . 51
6.3.2. Discontinuous Casting . . . . 51
6.3.3. Continuous Casting . . . . 51
6.3.4. Continuous Rod Casting and Rolling 51 6.4. Copper Powder . . . . 52
6.5. Copper Grades and Standardization 53 6.6. Quality Control and Analysis . . . . 54
7. Processing and Uses . . . . 55
7.1. Working Processes . . . . 55
7.2. Other Fabricating Methods . . . . . 56
7.3. Uses . . . . 57 8. Economic Aspects . . . . 58 9. Environmental Protection . . . . 60 10. Toxicology . . . . 61 11. References . . . . 62
1. Introduction
Copper [7440-50-8], the red metal, apart from gold the only metallic element with a color dif-ferent from a gray tone, has been known since the early days of the human race. It has always been one of the significant materials, and today it is the most frequently used heavy nonferrous metal. The utility of pure copper is based on its physi-cal and chemiphysi-cal properties, above all, its electri-cal and thermal conductivity (exceeded only by
silver), its outstanding ductility and thus excel-lent workability, and its corrosion resistance (a chemical behavior making it a half-noble metal). Its common alloys, particularly brass and bronze, are of great practical importance (→ Copper Alloys). Copper compounds and ores are distinguished by bright coloration, espe-cially reds, greens, and blues (→ Copper Com-pounds). Copper in soil is an essential trace ele-ment for most creatures, including humans.
Etymology. According to mythology, the
goddess Venus (or Aphrodite) was born on the Mediterranean island of Cyprus, formerly Kypros (Greek), where copper was exploited millennia before Christ. Therefore, in early times the Romans named it cyprium, later called cuprum. This name is the origin of copper and of the corresponding words in most Romance and Germanic languages, e.g., cobre (Spanish and Portuguese), cuivre (French), Kupfer (German), koper (Dutch), and koppar (Swedish).
History [21–24]. The first metals found by
Neolithic man were gold and copper, later sil-ver and meteoric iron. The earliest findings of copper are presumed to be nearly nine millen-nia old and came from the region near Konya in southern Anatolia (Turkey). Until recently the six-millennia-old copper implements from Iran (Tepe Sialk) were presumed to be the oldest. In the Old World, copper has been worked and used since approximately
7000 b.c. Anatolia
4000 b.c. Egypt, Mesopotamia, Palestine, Iran, and Turkestan
3000 b.c. Aegean, India 2600 b.c. Cyprus
2500 b.c. Iberia, Transcaucasia, and China 2200 b.c. Central Europe
2000 b.c. British Isles 1500 b.c. Scandinavia
Empirical experience over millennia has led to an astonishing knowledge of copper metallur-gical operations:
1) Native copper was hardened by hammering (cold working) and softened by moderate heating (annealing).
2) Heating to higher temperatures (charcoal and bellows) produced molten copper and made possible the founding into forms of stone, clay, and later metal.
3) Similar treatment of the conspicuously col-ored oxidized copper ores formed copper metal.
4) The same treatment of sulfide copper ores (chalcopyrite), however, did not result in cop-per metal, but in copcop-per matte (a sulfidic in-termediate). Not before 2000 b.c. did people succeed in converting the matte into copper by repeated roasting and smelting.
5) In early times, bronze (copper – tin alloy) was won from complex ores, the Bronze Age
be-ginning ca. 2800 b.c. At first, copper ores were smelted with tin ores; later, bronze was produced from metallic copper and tin. Brass (copper – zinc alloy) was known ca. 1000 b.c. and became widely used in the era of the Ro-man Empire.
In Roman times, most copper ore was mined in Spain (Rio Tinto) and Cyprus. With the fall of the Roman Empire, mining in Europe came to a virtual halt. In Germany (Saxony), mining activ-ities were not resumed until 920 a.d. During the Middle Ages, mining and winning of metals ex-panded from Germany over the rest of Europe. In the middle of the 16th century, the current knowledge of metals was compiled in a detailed publication [23] by Georgius Agricola, De Re Metallica (1556).
Independent of the Old World, the Indians of North America had formed utensils by working native copper long before the time of Christ, al-though the skills of smelting and casting were unknown to them. On the other hand, the skill of copper casting was known in Peru ca. 500 a.d., and in the 15th century the Incas knew how to win the metal from sulfide ores.
Around 1500, Germany was the world leader in copper production, and the Fugger family dominated world copper trade. By 1800, Eng-land had gained first place, processing ores from her own sources and foreign pits into metal. Near 1850, Chile became the most important producer of copper ores, and toward the end of the last century, the United States had taken the world lead in mining copper ores and in production of refined copper.
Technical development in the copper indus-try has made enormous progress in the last 120 years. The blast furnace, based on the oldest principle of copper production, was continually developed into more efficient units. Neverthe-less, after World War I, it was increasingly re-placed by the reverberatory furnace, first con-structed in the United States. Since the end of World War II, this furnace has been superseded slowly by the flash smelting furnace invented in Finland. Recently, several even more mod-ern methods, especially from Canada and Japan, have begun to compete with the older processes. An important development in producing crude metal was the application of the Besse-mer converter concept to copper metallurgy by
Manh`es and David (France, 1880): this prin-ciple is still the most widely used method for copper converting in the world.
Over time the requirements for copper pu-rity have become increasingly stringent. The in-vention and development of electrolysis by J. B. Elkington (England, 1865) and E. Wohlwill (Germany, 1876) made refining of high-purity copper possible.
In addition, the quantity of copper produced has increased immensely (Table 1). Since 1800, ca. 375× 106t of primary copper has been mined in the world, but of this only ca. 10× 106t was mined between 1800 and 1900.
Table 1. World mine production of copper (approximate, from
sev-eral sources)
Year Production, 103t Year Production, 103t
1700 9 1970 6400 1800 17 1975 7300 1850 57 1980 7900 1900 450 1985 8300 1950 2500 1990 9225 1955 3100 1995 10 050 1960 4200 1997 11 525 1965 5000
2. Physical Properties
Most properties of copper metal depend on the degree of purity and on the source of the metal. Variations in properties are caused by
1) Grade of copper, i.e., the oxygen con-tent: tough-pitch copper, deoxidized copper, oxygen-free copper
2) Content of native impurities (e.g., arsenic, bismuth) or remnants of additives (e.g., phos-phorus), which form solid solutions or sepa-rate phases at the grain boundaries
3) Thermal and mechanical pretreatment of the metal, which lead to states such as cast cop-per, hot-rolled copcop-per, cold-worked (hard) copper, annealed (soft) copper, and sintered copper
These property differences are caused by the defects in the crystal lattice. Two groups of prop-erties are to be distinguished:
1) Low dependence on crystal lattice defects, e.g., caloric and thermodynamic properties, magnetic behavior, and nuclear characteris-tics
2) High dependence on defects, e.g., electrical and thermal conductivity, plastic behavior, kinetic phenomena, and resistance to corro-sion
The variations in properties are caused ei-ther by physical lattice imperfections (disloca-tions, lattice voids, and interstitial atoms) or by chemical imperfections (substitutional solid so-lutions).
Atomic and Nuclear Properties. The
atomic number of copper is 29, and the atomic mass Ar is 63.546± 0.003 (IUPAC, 1983). Copper consists of two natural isotopes,63Cu (68.94 %) and65Cu (31.06%). There are also nine synthetic radioactive isotopes with atomic masses between 58 and 68, of which67Cu has the longest half-life, ca. 58.5 h.
Crystal Structure. At moderate pressures,
copper crystallizes from low temperatures up to its melting point in a cubic closest-packed (ccp) lattice, type A 1 (also F1 or Cu) with the co-ordination number 12. X-ray structure analysis yields the following dimensions (at 20◦C):
Lattice constant 0.36152 nm Minimum interatomic distance 0.2551 nm Atomic radius 0.1276nm Atomic volume 7.114 cm3/mol
There is also a high-pressure modification, which forms at ca. 400 MPa and 100◦C.
Density. The theoretical density at 20◦C, computed from lattice constant and atomic mass is 8.93 g/cm3. The international standard was fixed at 8.89 g/cm3in 1913 by the IEC (Interna-tional Electrotechnical Commission). The max-imum value for 99.999 % copper reaches nearly 8.96g/cm3.
The density of commercial copper depends on its composition, especially the oxygen con-tent, its mechanical and thermal pretreatment, and the temperature. At 20◦C, a wide range of values are found:
Cold-worked and annealed copper 8.89 – 8.93 g/cm3 Cast tough-pitch electrolytic copper 8.30 – 8.70 g/cm3 Cast oxygen-free electrolytic copper 8.85 – 8.93 g/cm3
The values for cold-worked copper are higher than those of castings because the castings have pores and gas cavities.
The density of copper is nearly a linear func-tion of temperature, with a discontinuity at the melting point: Temperature,◦C Density, g/cm3 solid copper 20 8.93 600 8.68 900 8.47 1 083 8.32 liquid copper 1 083 7.99 1 200 7.81
The solidification shrinkage is 4 %; the spe-cific volume at 20◦C is 0.112 cm3/g.
Mechanical Properties. Important
mechan-ical values are given in Table 2. High-purity cop-per is an extremely ductile metal. Cold working increases the hardness and tensile strength (hard or hard-worked copper); subsequent annealing eliminates the hardening and strengthening so that the original soft state can be reproduced (soft copper). The working processes are based on this behavior (Section 7.1). Impurities that form solid solutions of the substitutional type likewise increase hardness and tensile strength.
Table 2. Mechanical properties of copper at room temperature
Property Unit Anealed Cold-worked (soft) copper (hard) copper
Elastic modulus GPa 100 – 120 120 – 130 Shearing modulus GPa 40 – 45 45 – 50 Poisson’s ratio 0.35
Tensile strength MPa 200 – 250 300 – 360 Yield strength MPa 40 – 120 250 – 320 Elongation % 30 – 40 3 – 5 Brinell hardness (HB) 40 – 50 80 – 110 Vickers hardness (HV) 45 – 55 90 – 120 Scratch hardness ≈3
Pure copper has outstanding hot workability without hot brittleness, but the high-temperature strength is low. Detrimental impurities, those that decrease the strength at high temperatures, are principally lead, bismuth, antimony, sele-nium, tellurium, and sulfur. The concentration of oxides of such elements at the grain bound-aries during heating causes the embrittlement. However, such an effect can be desirable when free cutting is required. At subzero temperatures,
copper is a high-strength material without cold brittleness.
The changes in typical mechanical properties such as tensile strength, elongation, and hard-ness by heat treatment result from recrystalliza-tion [25]. The dependence of recrystallizarecrystalliza-tion temperature and grain size on the duration of heating, the amount of previous cold deforma-tion, and the degree of purity of copper can be determined from diagrams. The recrystallization temperature is ca. 140◦C for high-purity copper and is 200 – 300◦C for common types of copper. A low recrystallization temperature is usually advantageous, but higher values are required to maintain strength and hardness if the metal is heated during use.
Thermal Properties. Important thermal
val-ues are compiled in Table 3. The thermal con-ductivity of copper is the highest of all metals except silver.
Table 3. Thermal properties of copper
Property Unit Value
Melting point K 1356(1083◦C) Boiling point K 2868 (2595◦C) Heat of fusion J/g 210 Heat of vaporization J/g 4810 Vapor pressure (at mp) Pa 0.073 Specific heat capacity
at 293 K (20◦C)
and 100 kPa (1 bar) J g−1K−1 0.385 at 1230 K (957◦C)
and 100 kPa 0.494
Average specific heat 273 – 573 K (0 – 300◦C)
at 100 kPa (1 bar) J g−1K−1 0.411 273 – 1273 K (0 – 1000◦C)
at 100 kPa 0.437
Coefficient of linear thermal expansion 273 – 373 K (0 – 100◦C) K−1 16.9× 10−6 273 – 673 K (0 – 400◦C) 17.9× 10−6 between 273 and 1173 K (0 – 900◦C) 19.8× 10−6 Thermal conductivity at 293 K (20◦C) W m−1K−1 394
Electrical Properties. In practice, the most
important property of copper is its high elec-trical conductivity; among all metals only sil-ver is a better conductor. Both electrical ductivity and thermal conductivity are con-nected with the Wiedemann – Franz relation and show strong dependence on temperature (Ta-ble 4). The old American standard, 100 %
IACS (International Annealed Copper Stan-dard), corresponds to 58.0 MS/m at 20◦C, and it is still widely used in the United States. The corresponding electrical resistivity () is 1.7241×10−8Ω · cm, and the less usual resistiv-ity based on weight (densresistiv-ity of 8.89 g/cm3, IEC) is 0.1533Ω g m−1. The corresponding temper-ature coefficients are 0.0068 ×10−8Ω m K−1 (d/dT) and 0.00393 K−1 (−1d/dT). The theoretical conductivity at 20◦C is nearly 60.0 MS/m or 103.4 % IACS, and today com-mercial oxygen-free copper (e.g., Cu-OF) has a conductivity of 101 % IACS.
Table 4. Temperature dependence of thermal and electrical
conduc-tivity of copper Temperature Thermal conductivity, W m−1K Electrical conductivity, MS/m K ◦C 17 −2565 000 73 −200 574 460 113 −160 450 173 −100 435 110 273 0 398 60 293 20 394 58 373 100 385 44 473 200 381 34 573 300 377 27 973 700 338 15
The factors that increase the strength de-crease electrical conductivity: cold working and elements that form solid solutions. Elements that form oxidic compounds that separate at grain boundaries affect electrical properties only slightly. Copper may lose up to ca. 3 % of its con-ductivity by cold working; however, subsequent annealing restores the original value. There is a simple rule: the harder the copper, the lower is its conductivity.
Other Properties. High-purity copper is diamagnetic with a mass susceptibility of − 0.085 × 10−6 cm3/g at room temperature.
The dependence on temperature is small. How-ever, a very low content of iron can strongly af-fect the magnetic properties of copper.
The lower the frequency of light, the higher the reflectivity of copper. The color of a clean, solid surface of high-purity copper is typically salmon red.
The surface tension of molten copper is 11.25× 10−3N/cm at 1150◦C, and the dynamic viscosity is 3.5× 10−3Pa· s at 1100◦C.
Detailed physical-property information and data are to be found in the literature, particularly as tabular compilations [25–30].
3. Chemical Properties
In the Periodic Table copper is placed in the first transition series (period 4). It belongs to Group 11 and, together with silver and gold, forms the coinage metals. Its electron config-uration is [Ar] 3d104s1. Copper compounds are known in oxidation states ranging from +1 to +4, although the +2 (cupric) and the +1 (cuprous) are by far the most common. In aqueous solutions or below 800◦C, the +2 oxidation state is the most stable.
Copper(I) compounds such as CuCl and CuI are diamagnetic colorless materials, except for those whose color results from charge-transfer bands, for example, Cu2O. Cu+ions, [Ar] 3d10, are coordinated in a linear (two ligands) or tetra-hedral fashion (four ligands).
Copper(II) compounds such as CuSO4
· 5 H2O are paramagnetic blue or green
sub-stances, the color of which results from strong absorption bands in the region between 600 and 900 nm caused by d – d electron transfer pro-cesses. The Cu2+ion is a d9system and gener-ally sixfold coordinated in a distorted octahedral manner.
Copper(III) compounds are mostly diamag-netic. Cuprates like NaCuO2 can be obtained by heating the oxides in pure oxygen. In chem-istry only a few Cu3+complexes are known, but it appears that Cu3+plays an important role in biochemistry, especially with deprotonated pep-tides.
Copper(IV) compounds are not well known except for Cs2[CuF6].
Behavior in Air. Copper in dry air at room
temperature slowly develops a thin protective film of copper(I) oxide [1317-39-1]. On heating to a high temperature in the presence of oxygen, copper forms first copper(I) oxide, then cop-per(II) oxide [1317-38-0], both of which cover the metal as a loose scale.
In the atmosphere, the surface of copper oxi-dizes in the course of years to a mixture of green basic salts, the patina, which consists chiefly of the basic sulfate, with some basic carbonate. (In a marine atmosphere, there is also some ba-sic chloride.) Such covering layers protect the metal.
Behavior versus Diverse Substances.
While many substances scarcely react with copper under dry conditions, the rate of at-tack increases considerably in the presence of moisture. Copper has a high affinity for free halogens, molten sulfur or hydrogen sulfide.
Standard electron potentials of copper are as follows [31], [32]:
Potentials in standard (acid) solution:
Cu++ e−−→ Cu E0= 0.521 V Cu2++ 2 e−−→ Cu E0= 0.153 V
Potentials with complexing ligands:
[Cu(NH3)4]2++ 2 e−−→ Cu + 4 NH3 E0=− 0.11 V [Cu(CN)2]−+ e−−→ Cu + 2 CN−E0=− 0.43 V
As the standard electron potentials show, copper metal is stable to nonoxidizing acids like dilute sulfuric or hydrochloric acid, similar to the pre-cious metals. Dissolution of copper is possible in oxidizing acids such as nitric acid or hot con-centrated sulfuric acid. Also other redox systems such as iron(III) or copper(II) chloride solutions are suitable reagents for leaching copper in prac-tice.
Copper dissolves not only in oxidizing acids but also, for example, in ammonia or cyanide so-lutions in the presence of oxygen because stable complexes are formed. Also acetic acid together with oxygen or hydrogen peroxide attacks cop-per forming a green pigment called verdigris.
Free Cu+ions are not stable in aqueous solu-tion although Cu+ (3d10) has a filled d shell. Spontaneous disproportion into Cu2+ and Cu takes place.
2 Cu+−→ Cu2++ Cu E0= 0.37 V
K = [Cu2+]/[Cu+] = 106
The distorted octahedral coordination of six wa-ter molecules around the Cu2+ ion (d9) gives an additional stabilization energy (ligand-field effect). In aqueous solutions, Cu+ is only ex-istent in form of very stable complexes like [Cu(CN)2]−or in the presence of an excess of copper metal. Also, insoluble Cu+compounds such as cuprous oxide do not disproportionate in water.
By virtue of its large ionic radius and low electrical charge, the Cu+ ion is a soft acid. Therefore, the chemistry of copper in the oxida-tion state + 1 is predominated by reacoxida-tions with soft bases like iodine (CuI), sulfur (CuSCN), or unsaturated nitrogen ligands. In contrast, the chemistry of Cu2+, which is smaller and more highly charged, is dominated by hard lig-ands like oxygen ([Cu(H2O)6]2+) or nitrogen ([Cu(NH3)4]2+).
Copper is very stable in fresh water and also in sea water or alkali metal hydroxide so-lutions. Wastewater containing organic sulfur compounds can be corrosive to copper.
Figure 1. Pourbaix diagram for copper in highly dilute
aque-ous solution at normal temperature [35]
Corrosion [33], [34]. M. J. N. Pourbaix
has developed potential – pH equilibrium dia-grams for metals in dilute aqueous solutions [35]. Such graphs give a rough indication of the feasibility of electrochemical reactions. Figure 1 shows the behavior of copper at room temper-ature and atmospheric pressure. The Cu – H2O system contains three fields of different charac-ter:
1) Corrosion, in which the metal is attacked 2) Immunity, in which reaction is
thermody-namically impossible
3) Passivity, in which there is no reaction be-cause of kinetic phenomena
Gases and Copper [36], [37]. An exact
knowledge of the behavior of solid and liquid copper toward gases is important for production and use of the metal. With the exception of hy-drogen, [1333-74-0], the solubility of gases in molten copper follows Henry’s law: the solubil-ity is proportional to the partial pressure.
Oxygen [7782-44-7] dissolves in molten cop-per as copcop-per(I) oxide up to a concentration of 12.6wt % Cu2O (corresponding to 1.4 wt % O) (also see Fig. 32). Copper(I) oxide in solid cop-per forms a separate solid phase.
Sulfur dioxide [7446-09-5] dissolves in molten copper and reacts:
6 Cu + SO2 Cu2S + 2 Cu2O
Hydrogen is considerably soluble in liquid cop-per, and after solidification some remains dis-solved in the solid metal, although copper does not form a hydride. The solubility follows Siev-ert’s law, being proportional to the square root of the partial pressure because the H2molecules dissociate into H atoms on dissolution. Hydro-gen has high diffusibility because of its ex-tremely small atomic volume.
Hydrogen dissolved in oxygen-bearing cop-per reacts with copcop-per(I) oxide at high temcop-pera- tempera-tures to form steam:
Cu2O + 2 H −→←− 2 Cu + H2O(g)
Steam is not soluble in copper; therefore, it ei-ther escapes or forms micropores.
Nitrogen, carbon monoxide, and carbon diox-ide are practically insoluble in liquid or solid copper. Hydrocarbons generally do not react with copper. An exception is acetylene, which reacts at room temperature to form the highly explosive copper acetylides Cu2C2 and CuC2; therefore, acetylene gas cylinders must not be equipped with copper fittings.
4. Occurrence
In the upper part of the earth’s crust (16km deep), the average copper content is ca. 50 ppm.
It is thus about half as abundant as chromium, about twice as abundant as cobalt, and 26th in order of abundance of the elements in the acces-sible sphere of the earth. Table 5 shows average copper contents in natural materials.
Table 5. Typical copper contents of natural materials
Mineral Content, ppm
Basalt 85
Diorite 30
Granite 10
Sandstone 1
Copper ores (poor) 5 000 Copper ores (rich) 50 000 Native copper 950 000
Seawater 0.003
Deep-sea clays 200
Manganese nodules 10 000 Marine ore sludges 10 000 Earth’s crust (average) 50 Meteorites (average) 180
4.1. Copper Minerals
More than 200 minerals contain copper in defin-able amounts, but only about 20 are of impor-tance as copper ores (Table 6) or as semiprecious stones (turquoise and malachite). Copper is a typical chalcophilic element; therefore, its prin-cipal minerals are sulfides, mostly chalcopyrite, bornite, and chalcocite, often accompanied by pyrite, galena, or sphalerite.
Secondary minerals are formed in sulfide ore bodies near the earth’s surface in two stages. In the oxidation zone, oxygen-containing wa-ter forms copper oxides, basic salts (basic car-bonates and basic sulfates), and silicates. In the deeper cementation zone, copper-bearing solu-tions from these salts are transformed into sec-ondary copper sulfides (chalcocite and covel-lite) and even native copper of often high purity, e.g., in the Michigan copper district (Keweenaw Peninsula).
Other metallic elements frequently found in copper ores are iron, lead, zinc, antimony, and ar-senic; less common are selenium, tellurium, bis-muth, silver, and gold. Substantial enrichments sometimes occur in complex ores. For example, ores from Sudbury, Ontario, in Canada contain nickel and copper in nearly the same concentra-tions, as well as considerable amounts of plat-inum metals. The copper ores from Zaire and
Table 6. The most important copper minerals
Mineral Formula Copper, wt % Crystal system Density, g/cm3
Native copper Cu 99.92 cubic 8.9
Chalcocite Cu2S 79.9 orthorhombic 5.5 – 5.8
Digenite Cu9S5 78.0 cubic 5.6
Covellite CuS 66.5 hexagonal 4.7
Chalcopyrite CuFeS2 34.6tetragonal 4.1 – 4.3
Bornite Cu5FeS4/Cu3FeS3 55.5 – 69.7 tetragonal 4.9 – 5.3 Tennantite Cu12As4S13 42 – 52 cubic 4.4 – 4.8 Tetraedrite Cu12Sb4S13 30 – 45 cubic 4.6– 5.1
Enargite Cu3AsS4 48.4 orthorhombic 4.4 – 4.5
Bournonite CuPbSbS3 13.0 orthorhombic 5.7 – 5.9
Cuprite Cu2O 88.8 cubic 6.15
Tenorite CuO 79.9 monoclinic 6.4
Malachite CuCO3· Cu(OH)2 57.5 monoclinic 4.0 Azurite 2 CuCO3· Cu(OH)2 55.3 monoclinic 3.8 Chrysocolla CuSiO3· n H2O 30 – 36(amorphous) 1.9 – 2.3 Dioptase Cu6[Si6O18]· 6 H2O 40.3 rhombohedral 3.3 Brochantite CuSO4· 3 Cu(OH)2 56.2 monoclinic 4.0 Antlerite CuSO4· 2 Cu(OH)2 53.8 orthorhombic 3.9 Chalcanthite CuSO4· 5 H2O 25.5 triclinic 2.2 – 2.3 Atacamite CuCl2· 3 Cu(OH)2 59.5 orthorhombic 3.75
Zambia are useful sources of cobalt. Many por-phyry copper ores in America contain significant amounts of molybdenum and are the most im-portant single source of rhenium. The extraction of precious metals and other rare elements can be decisive for the profitability of copper mines, smelters, and refineries.
4.2. Origin of Copper Ores
Ore deposits are classified according to their mode of formation, but the origin of copper ores is geologically difficult to unravel, and some of the proposed origins are controversial. The classification distinguishes two main groups, the magmatic series and the sedimentary series.
Magmatic ore formation involves magma crystallization and comprises the following groups:
1) Liquid magmatic ore deposits originate by segregation of the molten mass so that the heavier sulfides (corresponding to matte) separate from the silicates (corresponding to slag) and form intrusive ore bodies. Ex-amples: Sudbury, Ontario; Norilsk, western Siberia.
2) Pegmatitic – pneumatolytic ore deposits de-velop during the cooling of magma to ca. 374◦C, the critical temperature of water. Ex-amples: Bisbee, Arizona; Cananea, Mexico.
3) Hydrothermal ore deposits result by further cooling of the hot, dilute metal-bearing solu-tions from ca. 350◦C downward, i.e., below the critical temperature of water. Such de-posits contain copper primarily as chalcopy-rite and satisfy ca. 50 % of the demand in the Western world. There are many exam-ples of different types of hydrothermal de-posits. Examples: Butte, Montana (gangue deposit); Tsumeb, Namibia (metasomatic deposit); Bingham Canyon, Utah; Chuquica-mata, Chile; Toquepala, Peru; Bougainville, Solomon Islands (impregnation deposits). Impregnation deposits are also called dis-seminated copper ores or porphyry copper ores (or simply porphyries) because of their fine particle size.
4) Exhalative sedimentary ore deposits origi-nate from submarine volcanic exhalations and thermal springs that enter into seawa-ter, and constitute a transitional type to sedi-mentary deposits. These ores are third in eco-nomic importance in the Western world. The actual formation of such sulfidic precipita-tions can be observed, for example, the ma-rine ore slimes in the Red Sea. Examples: Mount Isa, Queensland; Rio Tinto, Spain; Rammelsberg (Harz), Federal Republic of Germany.
The origin of sedimentary ore occurs in the exogenous cycle of rocks and may be subdivided into the following groups:
1) Arid sediments in sandstones and conglom-erates occur widely in the former Soviet Union as widespread continental zones of weathering with uneven mineralization. Ex-amples: Dsheskasgan, Kazakhstan; Ex´otica, Chile.
2) Partly metamorphized sedimentary ores in shales, marls, and dolomites form large strata-bound ore deposits, especially in the African copper belt, and represent the second most important source of copper to the West-ern world, as well as supplying nearly 75 % of its cobalt. Examples: Zaire (oxidation zone, oxidized ores6% Cu); Zambia (cementa-tion zone, secondary sulfide ores4 % Cu). 3) Marine precipitates have formed sedimen-tary ore deposits similar to the present phe-nomenon of sulfide precipitation by sulfur bacteria in the depths of the Black Sea. Ex-amples: Silesia (copper marl), Poland. 4) Deep-sea concretions lie in abundance on the
bottom of the oceans, especially the Pacific Ocean. These so-called manganese nodules could also be a source of copper in the future.
4.3. Copper Ore Deposits
Geologically, the main regions of copper ore de-posits are found in two formations: the Precam-brian shields and the Tertiary fold mountains and archipelagos. There are major producing coun-tries on every continent [38], [39].
North America: United States (Arizona, Utah, New Mexico, Montana, Nevada, and Michi-gan), Canada (Ontario, Quebec, British Columbia, and Manitoba), and Mexico (Sonora)
South America: Chile, Peru, Argentina, and Brazil
Africa: Zaire, Zambia, Zimbabwe, South Africa, and Namibia
Australia and Oceania: Queensland, Papua New Guinea
Asia: Former Soviet Union (Siberia, Kaza-khstan, and Uzbekistan), Japan, Philippines, Indonesia, India, Iran, and Turkey
Europe: Poland (Silesia), Yugoslavia, Portugal, Bulgaria, Sweden, and Finland
Antarctica may be an important source of copper ores in the foreseeable future.
4.4. Copper Resources
World primary copper reserves were estimated in 1991 at 552× 106t (Table 7) [41]. Reserves are identified resources and do not include undiscovered resources. With time the avail-able reserves have increased because of techni-cal progress in processing ores with low copper content and the discovery of new ore deposits [38], [39]. About 321× 106t were classified as minable copper ores under the technical and eco-nomical conditions at that time.
It is believed that a large potential of as-yet untouched deposits exist. Therefore, the poten-tially usable copper resources are estimated to be about three times as large as the reserve base. In addition to ores on land, there is an estimated amount of copper in deep-sea nodules of about 0.7× 109t.
Table 7. Copper ore reserves in 1991 [41]
Country Ore reserves, Percentage of 106t world reserves United States 90 16.3 Chile 120 21.7 Peru 31 5.6 Zambia 30 5.4 Zaire 30 6.3 Canada 23 4.2 Australia 21 3.8 Philippines 162.9 Indonesia 8 1.4 China 8 1.4 Poland 15 2.7 CIS 54 9.8 Other countries 10619.2
If one assumes that total production of pri-mary copper will remain stable, the identified reserves would last until 2040. With an increase in copper production of 2 – 3 %, which is more realistic, the duration of the known reserves will be reduced. However, these forecasts are quite unreliable because the growing use of secondary copper (recycling materials) and the discovery of new copper ore deposits are not considered.
4.5. Mining
Exploration, which is the search for ore deposits and their detailed investigation, is required to
ascertain the commercial feasibility of a poten-tial mine. The used geological, geochemical, or geophysical methods are complicated and ex-pensive. But often legal and political factors are more decisive for a new mine project than tech-nological or financial aspects. The methods used for exhausting the copper ore are depending to a great extent to the type of deposit. Important parameters are metal content of the ore and ge-ometry and depth of the ore body. These pa-rameters determine the working method. Under-ground and open-cast mining are the two basic techniques. As a generalization mining can be divided into the stages drilling, blasting, load-ing, and haulage of the ore.
Technological developments like the LHD technique (load, haul, dump) and the use of concrete for stowing have made mining more cost efficient. But underground mining is still more labor-intensive and expensive than open cast mining. Therefore, copper ores with an av-erage Cu content of 1 % or more or that con-tain other valuable metals in addition to copper (e.g., precious metals, nickel, cobalt) are mined underground, while ores with 0.5 % Cu, which represents a 100-fold enrichment of the average copper content in the earth’s crust, are mined by open-cast methods. The porphyry copper de-posits which are located near the surface can only be exploited by open-cast technique. They have become more important during the last decades. The first open pit was started at Bing-ham Canyon early this century. Today the ter-raced copper open pits are the largest ore mines in the world. They often cover more than one square kilometer and have working depths of several hundred meters. 100 000 t of crude ore are extracted per day. Today more than 50 % of the primary copper comes from open pits. Other less common methods for copper extraction are in situ leaching or ocean mining.
In situ leaching is a hydrometallurgical pro-cess in which copper is extracted by chemical dissolution in sulfuric acid. This method is suit-able for low-grade copper ore bodies for which customary mining operations would be uneco-nomical, as well as for the leaching of remenant ores from abandoned mines. In some cases, the ore body must be broken before leaching by blasting with explosives to increase the surface area for chemical reaction.
Ocean mining involves obtaining metallifer-ous raw materials from the deep oceanic zones. Two groups of substances are of interest: deep-sea nodules [43] and marine ore slimes [44]. The nodules (manganese nodules; see→ Manganese and Manganese Alloys, Chap. 9.) contain, in addition to iron oxides, ca. 25 % Mn, 1 % Ni, 0.35 % Co and 0.5 % (max. 1.4 %) Cu. Spe-cially equipped ships have collected and lifted these nodules from depths of 3000 – 5000 m; specific metallurgical and chemical methods for processing the nodules have been developed in pilot plants. Because of the extremely high ex-penses, large-scale operations of this type have not yet been undertaken. Marine ore slimes from the Red Sea (2200-m depth) average ca. 4 % Zn, 1 % Cu, and a little silver. Although methods for processing these slimes have been investigated, this resource is not now economically important.
5. Production
Over the years copper production methods have been subjected to a continual selection and im-provement process because of the need for (1) in-creased productivity through rationalization, (2) lower energy consumption, (3) increased envi-ronmental protection, (4) increased reliability of operation, and (5) improved safety in operation. During this development a number of tendencies have become apparent:
1) Decrease in the number of process steps 2) Preference for continuous processes over
batch processes 3) Autogenous operation
4) Use of oxygen or oxygen-enriched air 5) Tendency toward electrometallurgical
meth-ods
6) Increased energy concentration per unit of volume and time
7) Electronic automation, measurement, and control
8) Recovery of sulfur for sale or disposal 9) Recovery of valuable byproducts
The selection of a particular production method depends essentially on the type of avail-able raw materials, which is usually ore or con-centrate and on the conditions at the plant loca-tion.
About 80 % of primary copper production comes from low-grade or poor sulfide ores. Af-ter enrichment steps, the copper concentrates are usually treated by pyrometallurgical meth-ods. Generally, copper extraction follows the se-quence:
1) Beneficiation by froth flotation of ore to give copper concentrate
2) Optional partial roasting to obtain oxidized material or calcines
3) Two-stage pyrometallurgical extraction a) smelting concentrates to matte
b) converting matte by oxidation to crude (converter or blister) copper
4) Refining the crude copper, usually in two steps
a) pyrometallurgically to fire-refined copper b) electrolytically to high-purity electrolytic
copper
Figure 2 [13] illustrates the principal pro-cesses for extracting copper from sulfide ore.
About 15 %, with an increasing trend, of the primary copper originates from low-grade oxi-dized (oxide) or mixed (oxioxi-dized and sulfidic) ores. Such materials are generally treated by hy-drometallurgical methods.
The very few high-grade or rich copper ores still available can be processed by traditional smelting in a shaft furnace. This process is also used for recovering copper from secondary ma-terials, such as intermediate products, scrap, and wastes.
Figure 3 [13] illustrates the most important operations in copper extraction from various ox-idic ores.
5.1. Beneficiation
Sulfidic copper ores are too dilute for direct smelting. Smelting these materials would re-quire too much energy and very large furnace capacities. The copper ore coming from the mine (0.5 – 1 % Cu) must be concentrated by benefi-ciation. The valuable minerals like chalcopyrite are intergrown with gangue. Therefore, in the first step the lumpy ore is crushed and milled into fine particles (< 100 µm) to liberate the in-dividual mineral phases.
Typical equipment for crushing to about 20 cm are gyratory and cone crushers. Then wet
grinding in semi-autogenous rod or autogenous ball mills takes place. Size classification takes is performed in cyclones. In the next step of bene-ficiation, valuable minerals and gangue are sep-arated by froth flotation of the ore pulp, which exploits the different surface properties of the sulfidic copper ore and the gangue [46]. The hy-drophobic sulfide particles become attached to the air bubbles, which are stirred into the pulp, rise with them to the surface of the pulp, and are skimmed off as a froth of fine concentrate. The hydrophilic gangue minerals remain in the pulp. Organic reagents with sulfur-containing groups at their polar end, such as xanthates, are used as collectors in the flotation process. Additionally, modifiers like hydroxyl ions (pH adjustment) are used to select different sulfide minerals, for example, chalcopyrite and pyrite. Alcohols are used to stabilize the froth.
To obtain concentrates with highest possible purity and recovery rate, the flotation process usually consists of several stages which are con-trolled by expert systems. Various sensors for particle size, pH, density, and other properties are installed. Figure 4 gives an overview of a typical beneficiation process at a concentrator. In the first flotation stage, as much copper as possible is recovered in a rougher concentrate so that as little as possible goes to the tailings. To increase the copper recovery rate, often these tailings are leached with sulfuric acid. After re-grinding, the rough concentrate is cleaned in sev-eral flotation steps. After sedimentation in thick-eners and filtration in automatic filter presses or vacuum filters (ceramic disk) the typical cop-per concentrate contains 25 – 35 % Cu and about 8 % moisture. The moisture content of the con-centrate is a compromise between transporting water (cost) and avoiding dust generation dur-ing transport. Dewatered concentrates may heat spontaneously or even catch fire; therefore, ap-propriate precautions must be taken [47].
Copper concentrators typically treat up to 100 000 t of ore per day. They are located directly at the mines to achieve low transport costs. The copper recovery efficiency is over 90 %. About 95 % of the ore input goes into the tailings, which are stored in large dams near the mine and are used for water recycling to the flotation stages.
Separation of special copper ores such as those containing molybdenite or with high zinc
Figure 4. Overview of a typical beneficiation process at a concentrator
or lead content (Canada) is also possible by flota-tion methods. Flotaflota-tion of non-sulfide copper minerals is rare because these ores are mostly subjected to hydrometallurgical copper recov-ery, for example, heap leaching. In Zambia and Zaire, however, siliceous copper oxide ores are floated with fatty acid collectors, and dolomitic copper oxide ores are sulfidized with sodium hy-drogensulfide and then floated [48].
5.2. Roasting
Roasting can be used to prepare sulfide concen-trates for subsequent pyrometallurgical or hy-drometallurgical process. Partial roasting under oxidizing conditions may be carried out prior to smelting in reverberatory or electric furnaces. Complete oxidizing or sulfatizing roasting may be performed before leaching operations, espe-cially if other valuable metals such as cobalt are present in the concentrate. Reducing roasting may be carried out if copper concentrates with very high contents of impurities such as As are to be smelted.
However, roasting processes are today not very important for the copper extraction pro-cess. Only a few plants are still operating, for example Boliden’s R¨onnsk¨ar Smelter [211] and
Bor Smelter [212]. Since ca. 1975 combined roasting and matte smelting processes such as flash smelting have been favored because of their lower energy consumption and process gas han-dling advantages.
The roasting process has several effects: 1) Drying the concentrates
2) Oxidizing a part of the iron present 3) Controlling the sulfur content
4) Partially removing volatile impurities, espe-cially arsenic
5) Preheating the calcined feed with added fluxes, chiefly silica and limestone
Chemical Reactions. When the moist
con-centrates, which contain many impurity ele-ments, are heated, a multitude of chemical reac-tions occur. Because analysis of the many ther-modynamic equilibria is not practical, a few fun-damental systems are usually chosen. The most important is the ternary copper – oxygen – sulfur system (Fig. 5). The next most important system is the ternary iron – oxygen – sulfur system be-cause most sulfidic copper ores contain signifi-cant amounts of iron.
Initially, sulfides such as pyrite and chalcopy-rite decompose and generate sulfur vapor, which reacts with oxygen to form sulfur dioxide:
2 CuFeS2 −→ Cu2S + 2 FeS + S(g) S(g) + O2(g)−→ SO2(g)
The principal reactions, i.e., the formation of metal oxides, sulfur trioxide, and metal sulfates, are exothermic.
MS + 1.5 O2 MO + SO2 SO2+ 0.5 O2 SO3 MO + SO3 MSO4
In addition, there are secondary reactions, such as the formation of basic sulfates, ferrites (espe-cially magnetite), and silicates, the last provid-ing most of the slag in the subsequent smeltprovid-ing:
MO + MSO4 −→ MO · MSO4 MO + Fe2O3 −→ MFe2O4 FeO + Fe2O3−→ Fe3O4 MO + SiO2 −→ MSiO3
Representative reductive roasting reactions are:
FeS2 −→ FeS + S(g)
8 FeAsS−→ 4 FeAs + 4 FeS + As4S4(g)
Figure 5. Partial phase diagram of the ternary Cu – O – S
system [51]
Methods. There are several important
roast-ing methods; all involve oxidation at an elevated temperature, generally between 500 and 750◦C: 1) Partial (oxidizing) roasting is the conven-tional way of extracting copper from sul-fide concentrates. At 700 – 750◦C, the de-gree of roasting is determined by controlling the access of air. A predetermined amount of sulfur (30 – 50 % is removed, and only part of the iron sulfide is oxidized. The copper sulfide is relatively unchanged. These con-ditions are important for the formation of a suitable matte.
2) Total, or dead, roasting is occasionally used for complete oxidation of all sulfides for a subsequent reduction process or for special hydrometallurgical operations.
3) Sulfatizing roasting is carried out at 550 – 650◦C to form sulfates. This method yields calcines well-suited for hydrometal-lurgical treatment.
Roasters. Industrial roasting is done in two
types of roasters: fluidized-bed and multiple-hearth roasters. Both are continuously operated processes.
Oxidizing roasting is usually carried out in fluidized-bed roasters with short residence times in the range of seconds and high production rates up to 50 t of moist concentrate per hour. The oxidation reactions supply most of the required heat. About 30 – 50 % of the incoming sulfide is oxidized to SO2by using slightly oxygen en-riched air (up to 30 % O2). The off-gas is rich in SO2(6– 12 %) and suitable for conversion to sulfuric acid. The hot calcine is usually sent to reverberatory or electric furnace. The advantage of a roaster in front of a smelting furnace is the lower energy requirement of the smelting furnace and the higher matte grade. Examples of fluidized-bed roasters are Boliden R¨onnsk¨ar (Sweden), Bor Smelter (Yugoslavia) in front of pyrometallurgical copper extraction, and Cham-bishi (Zambia) ahead of a leaching plant.
Reductive roasting is usually carried out in multiple-hearth furnaces because of the long res-idence time (several hours) and the precise con-trol of temperature and gas composition on each hearth. These roasters have lower production rates and are fired by natural gas burners. An ex-ample for the reducing process is the treatment
of El Indio (Chile) concentrate, which contains 8 % As and about 22 % Cu. About 97 % of the ar-senic is removed during the roasting process at about 650 – 720◦C [213]. In former times the multiple-hearth roasters (Herreshoff furnaces) were widely used for the extraction of copper from concentrates.
5.3. Pyrometallurgical Principles
Smelting of unroasted or partially roasted sul-fide ore concentrates produces two immiscible molten phases: a heavier sulfide phase contain-ing most of the copper, the matte, and an oxide phase, the slag. In most copper extraction pro-cesses, matte is an intermediate.
5.3.1. Behavior of the Components
The most important equilibrium in copper matte smelting is that between the oxides and sulfides of copper and iron:
Cu2O + FeS Cu2S + FeO
Iron(II) oxide [1345-25-1] reacts with added sil-ica flux to form fayalite [13918-37-1], a ferrous silicate that is the main component of slag:
2 FeO + SiO2−→ Fe2SiO4
Liquid iron sulfide [1317-37-9] reduces higher iron oxides to iron(II) oxide:
3 Fe2O3+ FeS−→ 7 FeO + SO2(g) 3 Fe3O4+ FeS−→ 10 FeO + SO2(g)
The second reaction serves to remove magnetite [1309-38-2], which complicates furnace opera-tions because of its high melting point (1590◦C) [54].
The pyrometallurgical production of copper from sulfide ore concentrates may be considered as a rough separation of the three main elements as crude copper, iron(II) silicate slag, and sulfur dioxide. About 20 accompanying elements must be removed from the copper by subsequent refin-ing. Table 8 shows the distribution of important impurities among matte, slag, and flue dust. Pre-cious metals, such as silver, gold, platinum, and
palladium, collect almost entirely in the matte, whereas calcium, magnesium, and aluminum go into the slag.
Table 8. Average percentage distribution of the accompanying
ele-ments in copper smelting, p. 591[20]
Element Matte Slag Flue
dust Arsenic 35 55 10 Antimony 30 55 15 Bismuth 10 10 80 Selenium 40 – 60 Tellurium 40 – 60 Nickel 98 2 – Cobalt 95 5 – Lead 30 10 60 Zinc 40 50 10 Tin 10 50 40
Silver and gold 99 1 –
5.3.2. Matte
The ternary Cu – Fe – S system is discussed in detail in the literature [55–57]. Figure 6shows the composition of the pyrometallurgically portant copper mattes and the liquid-phase im-miscibility gap between matte and the metallic phase. In the liquid state, copper matte is essen-tially a homogeneous mixture of copper(I) and iron(II) sulfides: the pseudobinary Cu2S – FeS system.
Arsenides and antimonides are soluble in molten matte, but their solubility decreases with an increasing percentage of copper in the matte. Accordingly, when the arsenic concentration is high, a special phase, the so-called speiss, can separate. It is produced under reducing condi-tions in the blast or electric furnace, and its decomposition is complicated (→ Arsenic and Arsenic Compounds, → Antimony and Anti-mony Compounds).
Compositions of several copper
mat-tes are shown in the partial diagram
Cu2S – FeS – (Fe3O4+ FeO) (Fig. 7), which is a section of the quaternary Cu – Fe – O – S system. The density of solid copper mattes ranges bet-ween 4.8 g/cm3 (FeS) and 5.8 g/cm3 (Cu2S); liquid mattes have the following densities: 4.1 g/cm3(30 wt % Cu, 40 wt % Fe, 30 wt % S), 4.6g/cm3(50 wt % Cu, 24 wt % Fe, 26wt % S), and 5.2 g/cm3(80 wt % Cu, 20 wt % S).
Figure 6. Ternary Cu – Fe – S diagram showing copper
mat-tes and the miscibility gap [55]
Figure 7. Partial ternary Cu2S – FeS1.08– (Fe3O4+ FeO) diagram [58] showing mattes from various processes ◦ Reverberatory furnace; Flash smelting furnace; Elec-tric furnace;• Blast furnace; Converter
Table 9. Composition (wt %) of typical copper smelter slags [64]
Component Reverberatory Flash Noranda Peirce – Smith
furnace furnace reactor converter
Copper 0.4 – 0.61 – 2 8 – 10 2 – 8 Iron (total) 35 40 35 50 Silica 38 30 21 25 Magnetite 7 – 12 13 25 – 29 20 – 25 Ratio of Fe to SiO2 0.92 1.33 1.67 2.0 5.3.3. Slags
Slags from copper matte smelting contain 30 – 40 % iron in the form of oxides and about the same percentage of silica (SiO2), mostly as iron(II) silicate. Such slags can be consid-ered as complex oxides in the CaO – FeO – SiO2 system [59] or, because of the relatively low CaO content of most slags, in the partial dia-gram FeO – Fe2O3– SiO2[60] (Fig. 8). Ternary systems of these and other pertinent oxide sys-tems are found in the literature [61], [62]. Ta-ble 9 shows the general composition of some slag types. Important properties of copper slag systems are compiled in [63].
Figure 8. Ternary FeO – Fe2O3– SiO2diagram [60]
The objectives of matte smelting are to achieve a rapid, complete separationof matte
and slag, the two immiscible phases, and a min-imal copper content in the slag. The differing properties of slag and matte affect this separa-tion:
1) the low, narrow melting interval of slag
2) the low density of liquid slag (ca.
3.1 – 3.6g/cm3) and the difference in den-sity between molten matte and slag of ca. 1 g/cm3
3) the low viscosity and high surface tension of the slag
The ratio of the weight percent of copper in matte to that of copper in slag should be bet-ween 50 and 100. High matte grades generally cause high copper losses in slag. Such losses depend on the mass ratio of slag to copper pro-duced, which is usually between 2 and 3. Copper in slags occurs in various forms, including sus-pended matte, dissolved copper(I) sulfide, and slagged copper(I) oxide, partially as a silicate, which is typical of nonequilibrium processing.
Slags containing < 0.8 % copper are sold as products with properties similar to those of natural basalt (crystalline) or obsidian (amor-phous) or discarded as waste. When liquid slag is cooled slowly, it forms a dense, hard, crys-talline product that can be used as a large-size fill for riverbank protection or dike construction and as a medium-size crushed fill for roadbeds or railway ballast. Quick solidification, by pouring molten slag into water, gives amorphous granu-lated slag, an excellent abrasive that has partially supplanted quartz sand. Ground granulated slag is used as a trace element fertilizer because of its copper and other nonferrous metals.
Most of the newer copper smelting pro-cesses produce high-grade mattes, and the short residence time of the materials in the reac-tion chamber results in an incomplete approach to chemical equilibrium. Both factors lead to high amounts of copper in the slag, generally >1 wt %. Such slags must be treated by special methods for copper recovery (Section 5.5.1).
5.3.4. Oxidizing Smelting Processes
Nearly all pyrometallurgical copper processes are based on the principle of partial oxidation of the sulfide ore concentrates. Methods based
on the total oxidation of sulfide ores with subse-quent reduction to metal, avoiding the formation of copper matte, are used only rarely because of high fuel consumption, formation of copper-rich slags, and production of crude copper with a high level of impurities.
Prior to the 1960s, the most important way of producing copper was roasting sulfide concen-trates, smelting the calcines in reverberatory fur-naces, and converting the matte in Peirce – Smith converters. Since that time, the modern flash smelting process with subsequent converting has become predominant. Figure 9 shows the flow sheet of a modern copper smelter, from concen-trate to pure cathode copper, including the use of oxygen, recovery of waste heat, and environ-mental protection. Table 10 compares the impor-tant stages and processes of copper production, showing the range of the matte composition for each process.
Figure 9. Typical flow sheet for pyrometallurgical copper
production from ore concentrates [65]
5.3.5. Proposals
Numerous laboratory experiments and pilot-plant runs have been carried out to develop smelting methods based on elements other than
T able 10. Surv ey of p yrometallur gical processes for copper production [66]
oxygen. Two lines of development have dom-inated, reduction with hydrogen and chlorina-tion, but without leading to commercialization.
Reduction. A potential process involves the
reduction of chalcopyrite [70]:
2 CuFeS2+ 3 H2+ 3 CaO−→ Cu2S + 2 Fe + 3 CaS + 3 H2O CuFeS2+ 2 H2+ 2 CaO −→ Cu + Fe + 2 CaS + 2 H2O
The reduction by hydrogen is endothermic, but the overall reaction with calcium oxide is exothermic. A similar proposal [71] is based on the reaction of a metal sulfide with steam in the presence of calcium oxide:
Chlorination. The reactions of chalcopyrite
with chlorine are also of interest [72]: >500◦C:
CuFeS2+ 2 Cl2 −→ CuCl2+ FeCl2+ 2 S
<500◦C:
CuFeS2+ 3.5 Cl2−→ CuCl2+ FeCl3+ S2Cl2
followed by electrolysis of the molten copper (I) chloride.
The recently proposed thermoelectron chlo-rination process is another variation [73]:
CuFeS2+ 2 Cl2−→ CuCl + FeCl3+ 2 S
Electrolysis. Another approach, to avoid
converting, proposed the electrolysis of molten copper matte [74].
5.4. Traditional Bath Smelting
At the end of the Middle Ages, copper was duced by the German or Swedish smelting pro-cess that involved roasting reduction with up to seven process steps in small shaft furnaces.
Around 1700, reverberatory furnaces were con-structed in which the ore was processed by roast-ing, the so-called English or Welsh copper smelt-ing process, originally with ten process steps.
The large blast and reverberatory furnaces of the 1900s were derived from these principles. Later, the electric furnace for matte smelting was developed. Newer processes are the Isas-melt/Ausmelt/Csiromelt (furnace with vertical blowing lance), the Noranda and CMT/Teniente reactors (developed from converters), the Rus-sian Vanyukov, and the Chinese Bayia process.
5.4.1. Blast Furnace Smelting
The blast or shaft furnace is well-suited for smelting high-grade, lumpy copper ore. If only fine concentrates are available, they must first be agglomerated by briquetting, pelletizing, or sintering. Because of this additional step and its overall low efficiency, the blast furnace lost its importance for primary copper production and is currently used in only a few places, for example, Glogow in Poland.
Smaller types of blast furnace, however, are used to process such copper-containing materi-als as intermediate products (e.g., cement cop-per or copcop-per(I) oxide precipitates), reverts (e.g., converter slag, refining slag, or flue dusts), and copper-alloy scrap.
The construction of the furnace is basically related to that of the iron blast furnace, but there are considerable differences in design, es-pecially in size and shape: the copper blast fur-nace is lower and smaller, and its cross section is rectangular. Developments adopted from the steel industry include use of preheated air (hot blast), oxygen-enriched air, and injection of liq-uid fuels.
The furnace is charged with alternate addi-tions of mixture (copper-containing materials and accessory fluxes such as silica, limestone, and dolomite) and coke (which serves as both fuel and reducing agent). There are three zones in the furnace:
1) In the heating zone (the uppermost), water evaporates and less stable substances disso-ciate.
2) In the reduction zone, heterogeneous reac-tions between gases and the solid charge take place.
3) In the smelting zone, liquid phases react. The usual mode of operation is reduc-ing smeltreduc-ing, which yields two main prod-ucts. Sulfide ores are used to produce a matte (40 – 50 wt % Cu) and a disposal slag (ca. 0.5 wt % Cu). In contrast, oxide ores are processed directly to impure black copper (95 wt % Cu) and to a copper-rich slag. The two ore types can be smelted together to pro-duce matte and a slag with low copper content. Another product is top gas, which contains flue dust. Ores that contain high concentrations of arsenic and antimony also form speiss, which is difficult to decompose.
In Poland (KGHM Polska Meidz S.A. Smelters in Glogow I and Legnica) the blast-furnace technology is well adapted to Pol-ish copper concentrates, which contain about 20 – 30 % Cu like normal chalcopyrites but also 5 – 10 % of organic carbon and only 9 – 12 % S. Also these concentrates have relatively high lead (up to 2.5 %) and arsenic (up to 0.3 %) content. The organic carbon compounds provide about 40 – 60 % of the process energy; the rest is added by coke. The matte has about 58 – 63 % Cu and 3 – 6% Pb. The slag contains less than 0.5 % Cu. The off-gas from the blast furnaces (three in each plant) is mixed with the converter gases (Hoboken Converter). It contains 7 – 10 % SO2 and is sent to sulfuric acid production. The pro-duction figures are 80 000 t/a converter copper in Legnica, and 200 000 t/a in Glogow I Smelter [214].
5.4.2. Reverberatory Furnace Smelting
The reverberatory furnace dominated copper matte smelting for much of the 1900s, because it was an excellent process for smelting fine con-centrate from flotation. It is a fossil fuel fired hearth furnace for smelting concentrate and pro-ducing copper matte. The reverbs began to de-cline in the 1970s with the adoption of environ-mentally and energetically superior processes like flash smelting. Probably the last one was erected in 1976in Sar Chemesh, Iran. In 1980 about 100 reverbs were in operation, but in 1994 the number had decreased to about 25. As shown in Figure 10 it is a rectangular furnace up to 10 m wide and 35 m long with internal brick lin-ing. Throughputs of up to 1100 t/d concentrate
or a mixture of concentrate and calcine could be processed. The charge is passed into the fur-nace near the burners through the roof or lateral openings. As fossil fuel, pulverized coal, heavy oil, or natural gas is used. Normally the burner is located in the front wall of the furnace. The atmosphere is slightly oxidizing, and the maxi-mum flame temperature is up to 1500◦C. During the 1980s oxygen – fuel burners have been set in the roof to fire downwards directly on the top of the bath. This increases the smelting rate by up to 40 % and the energy efficiency to about 50 % [76], [215]. Another invention was the sprinkler burner for feeding concentrate, coal, and flux from the top of the reverb [77].
The back half of the furnace is the settling zone. A matte grade of between 35 and 60 % Cu is produced, depending on whether concentrate or calcine is fed. The slag has low copper con-tent (< 1 %). Normally, the converter slag is also fed back to the reverb. Sometimes problems with solid magnetite accretions in the furnace occur. The off-gas has a temperature of about 1250◦C. It is diluted by the combustion air and contains only about 1 – 2 % SO2. This is too low for efficient SO2 capture such as sulfuric acid production by the contact process. Improve-ments have been achieved by using higher oxy-gen enrichment of the burners, but the off-gas is still too dilute, and therefore this furnace is unsuitable for many regions in the world be-cause of environmental problems. Another dis-advantage is the very high energy consumption. The reverb has the highest energy consumption of all copper matte smelting processes. A good present example for operating reverbs is the Ilo Smelter in Peru (Table 11). There two reverbs smelt about 2000 t/d of concentrate, producing a matte containing 35 % copper for the subsequent Peirce – Smith converters. In the latest smelter enlargement one reverb was replaced by an Te-niente converter (CMT), and a sulfuric acid plant for partial SO2recovery was built.
5.4.3. Electric Furnace Smelting
In regions where relatively cheap electrical power was available, electric furnaces were built. The Scandinavian countries were the first to perform electric matte smelting: 1929 at Sulit-jelma, Norway; 1938 at Imatra, Finland, and
Figure 10. Schematic longitudinal and cross-section views through a reverberatory furnace [45]
1949 at R¨onnsk¨ar, Sweden, which is still op-erating.
Electric furnaces have the rectangular ground plan and the dimensions of larger reverbera-tory furnaces. Along the centerline, up to six S¨oderberg continuous self-baking electrodes are used with alternating current (Table 11). Table 11. Examples of traditional reverberatory and electric furnaces
Boliden R¨onnsk¨ar Smelter
Southern Peru Ilo Smelter
Type electric furnace reverbatory Inside dimensions
l× w × h, m
24× 6 × 3 36× 10 × 3.4 Electrodes/burners 68
Smelting capacity, t/d 950 (hot calcine) 1000 (wet concentrate) Matte grade, % Cu 51 35 Off-gas volume, m3/h (STP) 25 000 100 000 SO2content, % 4.5 1.2 Energy consumption 300 kW/t concentrate 194 kg oil/t
concentrate
The smelting process in such units is similar to the operation in reverbs, but the concentrate is usually dried and roasted before charging to in-crease the smelting capacity. The converter slag is returned to the furnace. The composition of matte and slag resembles that of the reverb prod-ucts, but the content of magnetite is lower. No fuel is burned, and the volume of waste gases and the quantity of flue dust are small. The SO2
content of the waste gas can be 10 %. The off-gas also contains carbon oxides from the consump-tion of the graphite electrodes.
A significant difference between the electric furnace and the reverb is the inversion of the tem-perature gradient in the furnace cavity. In the re-verb, the combustion gases have the highest tem-perature, whereas in electric matte smelting, the waste gases are ca. 500◦C cooler than the slag
phase, which is heated by the electric energy. Accordingly, in the electric furnace, cheap re-fractories are sufficient for lining the walls above the slag zone and the roof; only a common arch is required.
The temperatures of both molten phases de-pend on the submergence of the electrodes, and the required heat is controlled by the electrical power supplied. Heat transfer takes place chiefly by convection, which causes intense circulation in the molten bath.
The current is divided into two partial cir-cuits, through slag and through matte. The dif-ference in conductivity is great, slag : matte ra-tio of 1 : 102to 1 : 103; therefore, the depth of immersion of the electrodes into the liquid slag must be precisely controlled. As the electrodes immerse deeper into the slag, more current flows through the matte. If they touch the matte layer, a short circuit occurs. These considerations lead to an approximate relation between the electri-cal power input and the depth of the slag layer: 6000 kVA and 0.5 m, 30 000 kW and 1.0 m, and 50 000 kW and 1.5 m.
The smelting capacity of electrical furnaces is higher than that of reverbs.
Brixlegg Process. Lurgi developed and
prac-ticed at Brixlegg, Tirol, Austria, a modification of the old roasting reaction process. Nearly dead-roasted concentrates are reduced in batches by coal in a small special circular electric furnace (2500 kVA, 5-m diameter). The crude copper av-eraged only 95 wt % Cu, and the operation has been discontinued.
5.4.4. Isasmelt Furnace
The Isasmelt furnace and the closely related Ausmelt furnace were both developed in the 1980s on the basis of work conducted by CSIRO, Australia [216]. The process consists of a tall cylindrical furnace (small diameter but large height, internally brick lined) equipped with a vertical blowing lance. The lance is submerged into the bath (slag) and blows oxygen-enriched air and/or fossil fuel (natural gas). The moist concentrate and flux material is pelletized and fed through the roof of the furnace onto the bath. Most of the energy requirement comes from the reaction of the concentrate. The lance is made from stainless steel. It is cooled by swirling the air in the annulus between the pipes so that a protective layer freezes on that part of the lance which is immersed in the slag bath (0.3 m) [217]. The lance has to be replaced from time to time, for example, once a week. In the continuous Isas process, an emulsion of matte and slag is pro-duced, which is tapped periodically to a electric or fuel-fired settling furnace for matte/slag sep-aration. The off-gas is high in SO2. For cooling it passes through a waste-heat boiler.
Two Isasmelter for copper matte are running (Mount Isa, Australia; Cypus Miami, Arizona). Figure 11 shows the Cyprus furnace schemati-cally. About 90 t/h of copper concentrate is pro-cessed, and about 45 t/h of matte is produced (58 % Cu). The off-gas contains about 35 % SO2 directly behind the furnace and is then diluted by air to 8 – 9 %. The Isasmelt process has replaced the electric (reverberatory) furnace, which is now used for matte/slag settling. Smaller Isas-melt or AusIsas-melt furnaces have also been pro-posed or built for secondary copper smelters and primary smelters like in China. It is possible to run the process batchwise, smelting and con-verting the material in the furnace for producing blister copper. Also Isasmelt furnaces have been built for lead refining.
5.4.5. Noranda Process
The Noranda process was initially constructed as a continuous smelting and converting process which produced blister copper from copper con-centrate. The first reactor was built in 1973 at Horne smelter [105]. This direct-to-blister pro-cess operated from 1973 to 1975 [218]. It was
switched to high-grade copper matte smelting (as it now still operates) because of excessive impurity levels in the copper anodes and to in-crease the smelting rate.
Figure 11. Isasmelt furnace
The Noranda furnace [219] is a horizontal steel barrel with an inner brick lining (Figure 12. The diameter is about 5 m and it is about 20 m long. The process runs continuously. At one end of the reactor, pelletized wet concentrate, coal, flux, revert materials, and scrap are thrown into the furnace and on to the top of the molten bath by a high-speed slinger belt. Feeding fine con-centrate through the tuyeres is also possible. The feed material is absorbed and melted in the liquid matte/slag bath. On the side there are 20 – 40 tuy-eres in the cylindrical part of the vessel. Oxygen-enriched air (ca. 40 % O2) is blown through these tuyeres into the matte phase. The tuyeres have to be punched periodically, as in a Peirce – Smith converter. A layer of matte and slag must always be present in the furnace.
The oxidation reaction of the sulfides (mostly iron sulfide) and the added coal provide the pro-cess energy. Matte is tapped at the bottom of the vessel. Slag is periodically tapped at the
