Thermodynamic Consideration of Copper Matte Smelting Conditions with Respect to Minor Element Removal and Slag Valorization Options

Abstract

Valorization of slag from metallurgical industrial processes becomes ever more important as this may help optimizing and reducing the use of natural resources such as rock and sand. In view of this objective, metallurgical processes have to be analyzed in order to modify the slag composition so that it will meet the expected requirements of potential users of these resources. This paper gives the results of a thermodynamic analysis of the flash smelting process to understand the impact of operational parameters on slag chemistry and elemental partitioning in the process. It is shown that an increase of the matte grade and temperature leads to higher deportment of Pb, Zn and As to the slag. With increasing matte grade the volatilization of these elements decreases. Furthermore, the results indicate an increasing solubility of Cu and S in slag at higher temperatures.

Introduction

Partitioning of major and minor elements during flash smelting has been considered in primary copper production by thermodynamic modelling [1–11].

Although this information is very useful during process design and optimisation to estimate the partitioning, the existing literature however does not focus on consequences for valorisation of the slag in sufficient detail and with sufficient validity. Due to the high complexity of the interactions of the input parameters influencing the output of a copper smelting process, existing results of process models cannot be directly transferred to describe other processes with different input parameters. This means every case needs to be considered on its own.

The present paper describes a distinct case of a copper flash smelter using input data from Hamburg Smelter East. Thermodynamic calculations are conducted utilizing FactSage software. In this work it is evaluated the influence of process parameters such as temperature, matte grade and Fe/SiO₂ ratio on the partitioning of an existing copper smelting process. This will provide a basis to understand trends of the element partitioning when changing only one of this parameters each under real smelting conditions. The main purpose is to use the results to illustrate the relationship between copper smelting conditions and slag valorisation options to provide viable pathways to increase the valorisation of the slag.

Fundamentals

Copper Smelting

Understanding the main parameters influencing the result of the copper smelting process is important to reproduce an existing copper smelting process within a thermodynamic model. The following parameters could be identified from literature:

Matte grade: This parameter is defined as the Cu content in the matte in wt%. In practice matte grade is increased by increasing the oxygen partial pressure in the system leading to oxidation of less noble elements than copper and therefore removal of these elements from the matte. One of the main smelting reactions is given by:

${\text{FeS}}\left( {\text{l}} \right) + {\text{Cu}}_{2} {\text{O}}\left( {\text{l}} \right) = {\text{Cu}}_{2} {\text{S}}\left( {\text{l}} \right) + {\text{FeO}}\left( {\text{l}} \right)$

(1)

Practically the matte grade, is adjusted by the ratio of O₂ in the blast and concentrate feed rate, assuming constant slag composition with constant Fe/SiO₂ ratio and oxygen enrichment of the blast. For the reason that the matte grade determines the extent of oxidation of Fe and S in copper smelting, it also determines SO₂ evolution [12] and it influences the partitioning of the elements between the phases.

Fe/SiO₂ ratio: The ratio of Fe to silica in the slag is a key parameter to be adjusted during smelting. Adding silica is important for producing slag with proper physical and chemical properties and to allow the separation of matte and slag avoiding formation of a single oxysulfide liquid [13]. The role of silica therefore is first of all the adjustment of low viscosity and melting temperature as well as of control of magnetite content. As described in various literature, the Fe/SiO₂ is an important factor determining Cu losses in the slag [14–16]. In [15] Cu losses to the slag are described by the following equations based on the theory of dissolved Cu in oxidic and sulfidic form:

${\text{K}}=\frac{{{\text{a}}_{\text{Cu}{{\text{O}}_{0.5}}}}}{{{\text{a}}_{\text{Cu}}}\text{P}_{{{\text{O}}_{\text{2}}}}^{1/4}}$

(2)

${\text{K}}=\frac{{{\text{a}}_{\text{Cu}{{\text{S}}_{0.5}}}}}{{{\text{a}}_{\text{Cu}}}\text{P}_{{{\text{S}}_{\text{2}}}}^{1/4}}$

(3)

$\left( {\% {\text{Cu}}} \right)_{{{\text{diss}}.{\text{total}}}} = \left( {\% {\text{Cu}}} \right)_{\text{O}} + \left( {\% {\text{Cu}}} \right)_{\text{S}}$

(4)

Whereas a_M, and p_M characterize the activity and partial pressure of the species M, $\left( {\% {\text{Cu}}} \right)_{{{\text{diss}}.{\text{total}}}}$ describes wt% of dissolved Cu in slag and $\left( {\% {\text{Cu}}} \right)_{\text{O}}$ , and $\left( {\% {\text{Cu}}} \right)_{\text{S}}$ describe the wt% of dissolved Cu in sulfidic and oxidic form.

Other slag formers: Besides Fe and SiO₂ also other slag formers like CaO, MgO or Al₂O₃ affect the solubility of Cu, Pb, Zn and As in the slag. In [14] the iso activity curves of FeO, and the activity coefficient curves of ZnO, PbO and Cu₂O in the CaO–FeO–SiO₂ system are shown, from which it becomes clear that a change of the Ca/SiO₂ ratio leads also to a change of the activities of these components.
Oxygen enrichment: The oxygen enrichment of the incoming gas is important for controlling the total amount of gas entering a furnace. Examples in the industrial practice show that flash furnaces are operated at very different oxygen enrichment from 21% O₂ during start-up of Giuxi Smelter to 75% at Kennecott [17]. Furthermore this parameter is important to adjust the heat balance. Because the contained N₂ can be considered as inert it does not take part in the reaction but consumes reaction heat.
Matte entrapment: For the reason that copper losses occur not only as dissolved Cu in slag but also as entrained matte in slag, the Cu content in the bulk slag is given as:

$\left( {\% {\text{Cu}}} \right)_{\text{bulk}} = (\% {\text{Cu}})_{\text{entrained}} + \left( {\% {\text{Cu}}} \right)_{{{\text{diss}}.{\text{total}}}}$

(5)

Matte entrapment depends on turbulence in the furnace, slag fluidity, operating temperature [17] and other properties of slag and metal. Nakamura and Toguri [18] described the influence of this properties on settling of entrained matte. In industrial practice chemical analysis is mostly concluded by analysing the bulk composition of the slag. Also in older fundamental investigations bulk analysis of slag is given because x-ray spectroscopy or micro analysis were not available. When the partitioning is described using the bulk composition this includes the entrained matte of the slag. With thermodynamic calculations only the dissolved element contents in slag are considered. For evaluation of data from industrial measurements this difference needs to be considered.

Temperature: Due to the fact that temperature is one of the most relevant parameters affecting equilibrium by influencing the equilibrium constant of any reaction it is considered to be very important. It furthermore influences the heat balance of the reaction, physical properties of the phases like viscosity and physicochemical properties like solubility of the components in the phases.

Modelling of the flash smelting process is widely discussed in literature. Literature shows various ways to model the process and gives indications of how a model can be set up by using mathematical equations describing the thermodynamics of the system [1–3, 19]. Other literature describes the use of software packages for the modelling of copper smelting [4, 5, 20–26]. Together with thermodynamic databases these software packages form a powerful tool which is essential for today’s metallurgists.

Cu Losses in Slag

Silica is added in copper smelting as an additional part of the feed. As already described earlier the Fe/SiO₂ ratio in slag influences the equilibrium in the system affecting the partitioning.

Due to the fact that copper losses into the slag play a decisive role for the economics of copper production it was investigated in numerous publications [15, 27–31]. It was found that the Cu losses depend on the matte grade but show a wide spread, depending on the distinct situation, as illustrated in Fig. 1. This occurs due to the differences in the conditions of the process or experiments but also in the analytical method as described earlier. A trend of obviously increasing Cu content in slag was observed at matte grades of higher than 70 wt% Cu. In [15] the difference between Fe and SiO₂ saturation is described in terms of Cu solubility in the slag, giving an increase of Cu losses when matte grade decreases from 70 wt% Cu to 50 wt% Cu under Fe saturation. Minimizing the Cu losses is a generally important topic in Cu industry to improve the economics of the copper production processes. But additionally it is also important to minimize the Cu content in the slag to fulfill legal requirements and engineering standards for slag application.

../images/468727_1_En_39_Chapter/468727_1_En_39_Fig1_HTML.gif — Fig. 1
Solubility of copper at silica saturation: 1 – 1200 °C, Fe saturation [27]. 2 – 1300 °C, Fe saturation [27], 3 – Noranda furnace slag [28], 4 – 1300 °C, SiO₂ saturation [15], 5 – 1300 °C Fe saturation [15], 6 – 1250 °C [29], 7 – 1250 °C [30], 8 – 1250 ° C [31]

Current Situation of Legislation and Engineering Standards

The present paper gives information on the change of element partition under variation of major process parameters. However, to understand the influence of slag composition on valorization options it is necessary to know about the given restrictions for further use of this material. Copper slag in its potential applications has to fulfill several laws and regulations in environmental regards. Every copper production plant is normally designed to manage the impurity contents in the feed. Because the situation in terms of legislation, engineering standards and markets is different for each region or country the situation in Germany is discussed within this paper as an example.

Potential applications of copper slag together with an extract of relevant laws or regulations are illustrated in Fig. 2.

../images/468727_1_En_39_Chapter/468727_1_En_39_Fig2_HTML.gif — Fig. 2
Selected applications of Cu slag and their associated laws or regulations

Some of these regulations focus on maximum contents of heavy metals that are allowed in different types of waters. Non-ferrous elements from the slag can be dissolved in water by elution. If elution of the material is considered, various factors will influence the behavior of the material, like mineralogical composition, surface area or pH of the eluent. However, to consider the extremum, it can be concluded, that there will be no elution of an element, if it does not exist in the slag. In the technical conditions of hydraulic engineering, substitute building material ordinance and directive on surface water purity limits for the eluate in elution tests or in waters are given [33–35]. In federal directive on soil purity and substitute material building ordinance there are criteria given for composition of solids [32, 33]. However, in applications in which the material is not applied as part of the soil these regulations do not become relevant. The current situation of applying the regulations and legislation for the different use of the material is rather complex. The reason for this is that limits given in waters or eluates cannot be directly linked to the bulk composition of the slag for given reasons. Furthermore, the future situation of regulations might change and is therefore uncertain. In summary this means that no limits for the composition of the slag can be given. Nevertheless, it can be stated that possibilities of the marketability of the material now and in the future can be increased with decreasing the contents of non-ferrous elements like Cu, Zn, Pb and As.

Understanding the thermodynamics of the copper smelting by thermodynamic modelling helps to understand the influence of process parameters on the deportment of these critical elements to the slag. It will furthermore support the optimization of the industrial process to minimize the contents of these elements in the slag.

Methodology

The FactSage thermodynamic software with C153 database was used to perform the thermodynamic calculations. Improvements in copper smelter modelling by using this database are described in [23]. The average matte temperature at tapping was chosen as the basis of the calculation. In the calculation, flash smelting was assumed as a single step process. This means that the reactor was not separated into individual parts like reaction shaft and bath or even parts of those. This simplification was done to minimize the number of calculations and the observation of trends was considered to be of importance than the absolute numbers. The total gas pressure was assumed to be 1 atm. The number of degrees of freedom was fixed according to Gibbs phase rule and only one parameter was chosen as variable for the input of the calculation at the time. The calculation was adjusted in a way that the effect of the variable parameter on phase composition and partitioning of minor elements can be understood. The matte grade was adjusted by increasing the oxygen amount in the incoming gas. The partial pressure of SO₂ was fixed by regulating nitrogen and therefore oxygen enrichment in the incoming gas. Fe/SiO₂ ratio was adjusted as constant by balancing SiO₂ in the input of the calculation. The content of Al₂O₃, CaO and MgO in the slag was fixed by applying the same principle as for Fe/SiO₂ ratio.

The conditions of the calculations are based on input data of the Aurubis Hamburg smelter as given in literature. Tables 1, 2, 3 show the data used in the calculation. The concentrate and flue dust composition as well as the contents of Al₂O₃, CaO and MgO are based on the information given in [36]. As in feed was not given but is assumed to be 0.1 wt% in concentrate. All compositions of the materials were normalized to 100 wt% in sum. Flux is a variable in the input to adjust the Fe/SiO₂ ratio in the output of the calculation. For the reason that Al₂O₃, MgO, CaO and SiO₂ need to be adjusted in the input of the calculations, it was assumed that the concentrate does not contain these components. Only by this assumption the given composition of the output material can be adjusted as given in Table 2 in the whole range of calculations. The information of Table 3 as well as the value for Fe/SiO₂ ratio is based on reference [37], whereas for oxygen enrichment, a range of numbers was given in literature, but 60% was chosen as basis of the calculation. With this value for oxygen enrichment a $p_{{SO_{2} }} = 0.54$ atm was achieved as a result of the initial calculations. Because literature only gives an amount of converter dust added into the furnace but no composition of it, it was neglected for the calculation. Furthermore, input of hydrocarbons was not considered.

Table 1

Input material composition in wt% as given in literature

	Cu	Fe	S	SiO₂	Pb	Zn	As	O
Concentrate	29	23	28	0	0.3	1	0.1
Flux	–	–	–	100	–	–	–	–
FSF dust	29	17	12	4	3	4.5	0.6	24

Table 2

Output material composition (Al2O3, CaO, MgO and Cu in wt%) as given in literature

	Fe/SiO₂	Al₂O₃	CaO	MgO	Cu
slag	1.22	4	1	1	–
matte	–	–	–	–	63

Table 3

Throughput and process data as basis for calculation

Concentrate feed	3100 t/d
FSF dust recycle	230 t/d
Oxygen enrichment	60%
Temperature	1220 °C

Results and Discussion

The partitioning of minor elements is given depending on the matte grade as shown in Fig. 3. The calculation was done for the temperature of 1220 °C (as given in the base case from literature), 1250 °C and 1280 °C, respectively. Fe/SiO₂ ratio is kept constant for the whole range of calculations. The dashed lines in the diagram indicate the boundary between the phases. The difference between the value of the lower and the upper boundary on the y-axis for each phase indicates the partition of the element in the given phase. The partitioning of each species M to phase ph can be described by Eq. (6) whereas $P_{M}^{ph}$ defines the partition of species M in phase ph, $m_{M}^{ph}$ refers to the mass of species M in phase ph and $m_{M}^{fd}$ the mass of species M in the feed.

../images/468727_1_En_39_Chapter/468727_1_En_39_Fig3_HTML.gif — Fig. 3
Partitioning of As, Pb, and Zn depending on matte grade under the conditions as given in the graphs

$\varvec{P}_{\varvec{M}}^{{\varvec{ph}}} = \frac{{\varvec{m}_{\varvec{M}}^{{\varvec{ph}}} }}{{\varvec{m}_{\varvec{M}}^{{\varvec{fd}}} }}$

(6)

For all considered temperatures, partitioning of As, Pb and Zn in the matte decreases with increasing the matte grade (see Fig. 3). Volatilization of As is decreasing with increasing the matte grade, and for a given matte grade it is increasing with increasing the temperature. There is also a slight tendency of the decrease in Pb volatilization with the matte grade, and a trend of the increase in Pb volatilization with temperature. The volatilization of Zn is constant over the whole range of matte grades, but similar to As and Pb volatilization it increases with temperature for a given matte grade. As partition in the gas phase can reach 20–40% (depending on the conditions, such as temperature and matte grade), which is highest amongst the considered elements. Less than 10% Zn and 20% Pb can be achieved in the gas phase.

According to the calculated results in Fig. 3, it can be concluded that there is:

Lower volatilization of the minor elements with increasing matte grade.
Higher volatilization with increasing temperature.
Higher slagging with matte grade.

These observations are in a good agreement with the results in the previous studies [6, 8, 10, 38, 39]. In reference [10] the volatilization of As and Pb is considered in a copper smelting fayalite slag system. However, the oxygen enrichment in the present system was 60% which is higher than that in reference [10], where it was 21–42%. This means that the gas volume in this reference is higher, leading to higher volatilization. In this reference activity coefficients of the elements in the slag as well as distribution coefficients between matte and slag are based on single parameter e.g. oxygen partial pressure or temperature. Therefore, the calculated activity coefficient can be considered as independent on the slag composition. This simplification will cause a deviation of the results with the present paper.

In terms of the further use of the slag it is relevant to understand not only the change of partitioning but also of slag composition, which is illustrated in Fig. 4. It that the content of Zn decreases and the content of As increases slightly. There is a significant increase of the Pb content in the slag with the matte grade. To explain that it is necessary to consider Fig. 3. Comparing the elements partition for 1220 °C matte grade of [Cu] = 50 wt% and [Cu] = 75 wt% it can be seen that the partition of As, Pb and Zn in slag is increased approximately by the factors 3, 2 and 5. The slag volume in that range matte grade increases by factor 2.2. Because the factor for Zn is lower, the content of Zn in the slag decreases, because the factor for Pb is significantly higher it increases in a relevant way. This means especially Pb and As have to be considered carefully in terms of valorisation of the material. The effect of temperature cannot be considered as significant for Pb and Zn but becomes more important for considering As.

../images/468727_1_En_39_Chapter/468727_1_En_39_Fig4_HTML.gif — Fig. 4
Slag composition depending on matte grade under the conditions given in the graph

Figure 5a and b show the Cu and S content in the slag as a function of the matte grade, respectively. For matte between 50 and 65 wt% the Cu content in the slag is not affected much. However, with increasing temperature also an increase of the content of dissolved Cu below a matte grade of [Cu] = 55 wt% becomes more obvious. The Cu content in slag increases significantly up to 75 wt% Cu in the matte. There is a big impact of the matte grade on the dissolved S in slag. The S content in slag decreases with increasing the matte grade.

../images/468727_1_En_39_Chapter/468727_1_En_39_Fig5_HTML.gif — Fig. 5
Cu and S content in the slag depending on matte grade under the conditions as given in the graphs

It is clear that for a given matte grade, the content of the dissolved S and Cu in slag increases with temperature. This could be an indication of the larger solubility of matte in slag at higher temperature. However, influence of the matte grade on the contents of both elements in slag is different, suggesting that other reaction mechanisms must be involved.

According to reaction (7), the decrease of the S content in the slag can be expected due to an increase of the oxygen partial pressure and therefore decrease of the S partial pressure in the system with increasing the matte grade at constant SO₂ partial pressure in the system.

$0.5\,{\text{S}}_{2} \left( {\text{g}} \right) + {\text{O}}_{2} \left( {\text{g}} \right) = {\text{SO}}_{2} \left( {\text{g}} \right)$

(7)

In [40] the dependence of S on the matte grade is given based on industrial measurements, which confirm the results of the model in the present paper with approx. 0.7% S at 50% Cu in matte decreasing to 0.1% at 75% Cu in matte. This implies that the model can be directly used for predictions of the S content in the slag. The dissolved Cu in slag increases strongly at higher matte grade. This is in line with the data of references [15, 28, 29] and [31] in Fig. 1. This can be explained through Eqs. (2) and (3). At lower matte grade the S partial pressure is higher, leading to higher solubility of Cu₂S in the system. This is confirmed by the results of the calculations showing higher contents of both elements at lower matte grade. As matte grade increases S is removed from the slag but Cu content increases especially when matte grade is larger than 70%, indicating that solubility of Cu₂O increases due to the increasing oxygen partial pressure (see Eq. (2)).

Conclusion

It was demonstrated that with the help of thermodynamic calculation the elements partitioning in copper smelting can be estimated. The results can be used to explore optimum process conditions in terms of Cu losses and minor element removal to the slag. It shows that the process parameters, i.e., matte grade and temperature have major influence on the partitioning of minor elements in the system and on the Cu and S content in the slag. With increasing matte grade the volatilization of the investigated minor elements is decreased. The deportment of these elements to the slag is increased. The S content in slag is decreased and the Cu content in slag increased when matte grade is larger than 70%. Increase of temperature increases the solubility of Cu and S in slag and increases the volatilization of minor elements. This illustrates that when changing one of the parameters the elements Zn, Pb and As behave opposingly to the elements Cu and S in terms of the objective to minimize the partitioning to the slag.

However, for the use of the material the composition plays a very important role. It could be shown that especially for Pb a significant increase in slag composition has to be expected when raising the matte grade. The increase of the Cu content in slag can especially be observed above a matte grade of [Cu] = 65 w.t%. Furthermore a significant decrease of the S content in the slag can be observed with increasing the matte grade. Low smelting temperatures are beneficial to minimize the Cu and S content in the slag. It was shown that application options for slag depend on its composition linked to given regulations. For this reason a compromise has to be found in order to achieve compliant Cu, S, As, Pb and Zn for the targeted application of the slag and its corresponding restrictions.

References

1.
Nagamori M, Mackey PJ (1978) Thermodynamics of copper matte converting: part II. Metall Trans B 9:567–579Crossref
2.
Chaubal PC, Sohn HY, George DB, Bailey LK (1989) Mathematical modeling of minor element behavior during flash smelting of copper concentrates and flash converting of copper mattes. Metall Trans B 20:39–51Crossref
3.
Seo KW, Sohn HY (1991) Mathematical modeling of sulfide flash smelting process: part III. Volatilization of minor elements. Metall Trans B 22(6):791–799Crossref
4.
Chen C, Zhang L, Jahanshahi S (2010) Thermodynamic modeling of arsenic in copper smelting processes. Metall Mater Trans B 41:1175–1185Crossref
5.
Chen C, Zhang L, Wright S, Jahanshahi S (2006) Thermodynamic modelling of minor elements in copper smelting processes. In: Sohn international symposium advanced processing of metals and materials, vol 1, pp 335–348
6.
Tan P, Zhang C (1996) Effects of temperature on distribution behaviors of minor elements in copper flash smelting—computer simulation. Trans Nonferrous Met Soc China 6(4):38–41
7.
Tan P, Zhang C (1997) Computer model of copper smelting process and distribution behaviours of accessory elements. J Cent South Univ Technol 4(1):36–41Crossref
8.
Tan P, Zhang C (1997) Influence of ratio of iron-to-silicon in slag on behaviors of associated elements during copper smelting. Shanghai Non-ferrous Metals 18(4)
9.
Kim HG, Sohn HY (1997) Minor-element behavior and iron partition during the slag cleaning of copper converter slag under reducing conditions. Can Metall Q 26(1): 31–37
10.
Kim HG, Sohn HY (1996) Thermodynamic modelling of minor-element behaviour in in-bath copper smelting and converting with calcium ferrite slag. Trans Instr Min Metall 105:151–163
11.
Kim HG, Sohn HY (1991) Prediction of minor-element behavior in copper smelting and converting with submerged oxygen injection. EPD Congr 91:437–465
12.
Davenport W, Jones D, King MJ, Partelpoeg EH (2003) Flash smelting: analysis, control and optimization, 2nd edn. TMS, Warrendale, pp 1–30
13.
Yazawa A, Kameda A (1953) Copper smelting. I. Partial liquidus diagram for FeS-FeO-SiO₂ system. Technol Rep Tohoku Univ 16:40–58
14.
Yamaguchi K, Ueda S, Takeda Y (2005) Phase equilibrium and thermodynamic properties of SiO₂-CaO-FeOx slags for copper smelting—research achievements of Professor Yoichi Takeda. Scand J Metall:164–174
15.
Takeda Y (2001) Copper solubility in SiO₂-CaO-FeOx slag equilibrated with matte. High Temp Mater Process (Lond) 20(3–4):279–284
16.
Chen M, Cui Z, Zhao B (2015) Slag chemistry of bottom blown copper smelting furnace at Dongying Fangyuan. In: 6th international symposium on high-temperature metallurgical proceeding, pp 257–264
17.
Larouche P (2001) Minor elements in copper smelting and electrorefining. MSc thesis, McGill Universtiy, Montreal
18.
Nakamura T, Toguri JM (1991) Interfacial phenomena in copper smelting processes. Copper 91(9):537–551
19.
Shimpo R, Watanabe S, Goto S, Ogawa O (1983) An application of equilibrium calculations to the copper smelting operation. Adv Sulfide Smelt:295–316
20.
Bale CW, Belisle E, Chartrand P, Decterov SA, Eriksson G, Gheribi A, Hack K, Jung I-H, Melancon J, Pelton AD, Petersen S, Robelin C (2014) Recent developments in FactSage thermochemical software and databases. In: Celebrating the megascale: proceedings of the extraction and processing division symposium on pyrometallurgy in honor of David G. C. Robertson, pp 141–148
21.
Bale CW, Chartrand SA, Degterov SA, Eriksson G, Hack K, Ben Mahfoud R, Melancon J, Pelton AD, Petersen S (2012) FactSage thermochemical software and databases. Calphad 26(189–228):2002
22.
Degterov S, Pelton A (1999) A thermodyanmic database for copper smelting and converting. Metall Mater Trans B 30:661–669Crossref
23.
Jak E, Hidayat T, Shishin D, Mehrjardi AF, Chen J, Hayes P (2017) Experimental and modelling research in support of energy savings and improved productivity in non-ferrous metal production and recycling. Proc EMC 2017:231–250
24.
Cardona N, Coursol P, Mackey PJ, Parra R (2011) Physical chemistry of copper smelting slags and copper losses at the Paipote smelter part 1—thermodynamic modelling. Can Metall Q 50:318–329Crossref
25.
Tan P (2004) CuModel—a thermodynamic model and computer program of copper smelting and converting processes and its industrial applications. EPD Congr 2004:411–422
26.
Gisby J, Taskinen P, Pihlasalo J, Li Z, Tyrer M, Pearce J, Avarmaa K, Björklund P, Davies H, Korpi M, Martin S, Pesonen L, Robinson J (2017) MTDATA and the prediction of phase equilibria in oxide systems: 30 years of industrial collaboration. Metall Mater Trans B 48:91–98Crossref
27.
Yazawa A, Kameda M (1954) Fundamental studies on copper smelting (II) solubilites of constituents of matte in slag. Tech Rep Tohoku Univ 19:1–22
28.
Spira P, Themelis J (1969) The solubility of copper in slags. J Metals:35–42
29.
Imris I (1975) Thermodynamics of reducing copper losses in slags. Copper Metall Pract Theory Pap Meet:18–22
30.
Jalkanen KH (1981) Copper and sulphur solubilities in silica saturated iron silicate slags from copper mattes. Scand J Metals 10:77–184
31.
Geveci A, Rosenqvist T (1973) Equilibrium relations between liquid copper, iron-copper matte and iron silicate slag at 1250 °C. Miner Process Extr Metall 82C:193–201
32.
Verordnung über Deponien und Langzeitlager-Deponieverordnung (DepV) (2009)
33.
Bundesministeriums für Umwelt, Naturschutz, Bau und Reaktorsicherheit. http://www.bmub.bund.de/fileadmin/Daten_BMU/Download_PDF/Abfallwirtschaft/mantelv_entwurf_bf.pdf. Accessed 27 Oct 2017
34.
Verordnung zum Schutz der Oberflächengewässer—Oberflächengewässerverordnung (OGewV) (2016)
35.
Technische Lieferbedingungen für Wasserbausteine (2003)
36.
Davennport WG, Partelpoeg EH (1989) Flash smelting analysis control and optimization. Pergamon Press
37.
Schlesinger ME, King MJ, Sole KC, Davenport WG (2011) Extractive metallurgy of copper, 5th edn. Elsevier
38.
Tan P, Zhang C (1998) Behaviours of accessory elements in copper pryometallurgy. Trans Nonferrous Met Soc China 8(1):114–119
39.
Tan P, Zhang C (1997) Modeling of accessory element distribution in copper smelting process. Scand J Metall 26:115–122
40.
Mackey PJ (1982) The physical chemistry of copper smelting slags—a review. Can Metall Q 21(3):221–260Crossref