Argentinean Copper Concentrates: Structural Aspects and Thermal Behaviour

In Argentina, there are many sources of copper concentrates. Some of them are currently in operation, while others are in the exploration stage. All copper concentrates produced are exported to other countries for copper refinement and to create various finished products. It is desirable that in the near future, these copper concentrates be processed in an Argentinean industrial plant. The aim of this paper was to present the results of a characterisation study carried out on five different copper concentrate samples. The thermal decomposition of the copper concentrates was determined by differential thermal analysis and thermogravimetry (DTA TG). The information was correlated with the chemical composition and the mineralogical phases of the samples identified by X-ray diffraction. A melting test at temperatures of up to 1300 ̊C was performed to complete the study of the concentrate’s behaviour during heating. After the test, all of the samples were observed by light and electronic scanning microscopy to identify the different phases generated under high-temperature conditions.


Introduction
The copper market is undoubtedly one of the most important metallic markets in the world.This statement is based on the significantly increased global demand for this metal and its alloys in recent years.
Argentina has large reserves of copper ores and constitutes a strong supplier of concentrates in our region [1][2][3][4].The ores selected for this study contain between 0.5% -0.8% copper and are free of impurities such as Sb, As, Te and Se.Currently, concentrates obtained by flotation operations are exported to other countries to produce final copper products such as wires, pipes and sheets [5].It is important to promote the installation of a copper pyrometallurgy plant in our country to allow for the ability to process concentrates and create final products for Argentina´s domestic market.Copper industrial processes include melting, conversion and refining operations [6].To design the pyrometallurgy process, it is necessary to completely characterise the concentrates.Thus, it is necessary to obtain information about the current phases that exist in concentrate particles and the type of transformations and reactions that occur under processing conditions [7].
Winkel et al. [8] described the importance of enhancing the knowledge about the high-temperature behaviour of copper iron sulphides, as well as the volatile impurities they contain.This information is essential to develop new processes of extractive metallurgy.
In this paper, information regarding the thermal behaviour of five samples of Argentinean copper concentrates is presented.The results obtained by DTA and TG were correlated with information determined by X-ray diffraction (XRD) and a microscopy study.Structural analyses of the concentrate samples were performed after the melting tests.The test products were examined by electron microscopy (SEM), and the phases were identified by XRD.The authors describe the mechanisms, such

Samples
Five samples of copper concentrates obtained by Rougher flotation were selected for this study.The copper ores were obtained from sources located in northern Argentine.

Methods
The concentrate characterisation was carried out.The chemical composition of each sample was determined by acid attack and atomic absorption spectrometry using a Perkin Elmer AA110 instrument.The crystal phases were identified by X-ray diffraction (XRD) at room temperature using a Philips X'Pert diffractometer.The sulphide quantification was conducted by applying the ASTM standard C25 [13].
A microscopy study performed using an Olympus GX51 light microscope and applying a LECO IA32 analysis system, which allowed for the observation of the types of particles and their morphological characteristics.
Another objective of this paper was to evaluate the behaviour of the concentrates at high temperatures.To obtain information about the reactions that occur duringheating, melting tests were carried out at temperatures of up to 1300˚C in air.Samples were melted in a porcelain crucible using an electric furnace.All of the melted samples were prepared for microscopy observation.Their structure was studied using light and scanning electron microscopy (SEM).Finally, the results were correlated with the DTA-TG results.

Chemical Composition
The chemical composition of the five concentrates samples is presented in Table 1.

Mineral Phases
The X-ray diffraction results provide information regarding the crystalline phases present in the five samples studied.The major minerals present in the concentrates were chalcopyrite (CuFeS 2 ), pyrite (Cu 2 S) and iron sulphide (FeS).Nevertheless, other minerals such as iron bisulphide (FeS 2 ), geerite (Cu 8 S 5 ), enargite (Cu 3 AsS 4 ) and dicopper zinc silicon tetrasulphide (Cu 2 Zn SiS 4 ), among others, were identified.However, sample E was observed to contain a lower chalcopyrite content and higher pyrite content.In sample B, geerite (Cu 8 S 5 ), enargite (Cu 3 AsS 4 ) and dicopper zinc silicon tetrasulphide (Cu 2 Zn SiS 4 ) were identified, and in sample C, PbS was identified.Traces of cassiterite (SnO 2 ) were identified in samples A and D. Figures 1-5 show the diffractograms of all of the samples, specifically the principal peaks corresponding to the crystal phases observed.Copyright © 2013 SciRes.IJNM

Thermal Analysis Tests
DTA and TG tests were carried out at a heating rate of 10˚C/s using a JENCK instrument.In this manner, the mechanisms of oxidation during heating up to 1000˚C were characterised for all of the copper concentrate samples.For each sample, the first exothermic peak temperature, T in , was determined.This temperature establishes the beginning of the oxidisation reaction.The  exothermic heat, Q ex , was determined by the area below the DTA curve.These values represent the total heat that developed during all of the reactions.Both values are very useful for understanding the flash-melting process [11,12,14,15].Figure 6 shows the DTA curves of the five samples.The T in and Q ex values calculated by numerical integration are presented in Table 2. Notably, both values are related to the oxygen content of the copper, which in this case, are constant.
The thermogravimetric curves (TG) (Figure 7) allow for the observation of mass changes and losses in the concentrate samples during heating.

Melting Test
Samples of each copper concentrate were melted in porcelain crucibles at 1200˚C at a heating rate of 5˚C/min in air.The structure of the solidified product of each sample was studied by light and sc ng electron microscopy anni   EM).In all cases, three different layers were recog-ect to the bottom-layer opper bands and globular pa (S nised: a greyish layer with a fine dispersion of white particles (located in the bottom of the crucible), a white (intermediate) layer that contained a large number of dendritic crystals and extensive porosity and finally a top grey layer composed of spherical and irregular white particles.Figure 8 shows a vertical cross section of the solidified layers in sample A. particles of a larger size with resp Notably, in the intermediate white layer, native copper bands (Figure 9) and white dendrites were observed.A higher proportion of native c rticles was observed in sample B. In the grey top layer, white crystals with different morphologies, irregular, needle-shaped and dendritic, were observed.
In samples B, C, D and E, the bottom and top grey layers were observed to be thicker than those in sample A and to contain a low proportion of white particles.However, in all of the samples, the intermediate white layer was observed to contain metallic copper bands.
The phases were examined by SEM in the three layers.Elemental mapping revealed the elemental distribution in each layer.Table 3 shows the elemental distribution ob- The botto p phases with higher ir terface in contact with the intermediate white layer.Also, traces of other elements such as Mg and Cr in the silicoaluminate matrix were detected.
The intermediate white layer contains metallic copper bands, white dendrites or irregular crystals of (Cu, Fe)S, and the matrix has a high iron sulphide ( The white and the grey top layers were observed to be porous.Sample D showed the highest porosity among the samples studied.
The top grey layer matrix consists of calcium silicoaluminate and copper sulphide particles (with dendritic and globular morp entified.Irregular copper particles isolated in the layer and in the microcracks were also observed. If we consider copper, iron and sulphur to be the main elements and approximate their content to be 100%, it is possible to plot on a ternary diagram the s sition identified in the intermediate white layer and in the grey top layer.For simplicity, the plotted compositions correspond to samples A and D (Figure 10).
These two samples were selected because of the phases differences observed.Samples B, C and E contain phases similar to those observed in sample A but in rent proportions.Notably, the sulphide phases developed under the test conditions (1200˚C) show copper contents up to 52%.

Discussion
The chemical compos sample B possess lower content of also exhibited the lowest S/Cu ratio (1, 58).Based on the DTA results, the correlation between the T in and Q ex values with respect to the law of copper concentration (C lynomial equations, (Equations ( 1) and ( 2)) for which R = 0.9946.
The T in values vary with mineralogical com and respect the following order of the samples: B < D < E. As observed experimentally, sample B has an in erally as cant loss of mass during heating between room te position A < C < termediate initial oxidation temperature.A higher content of pyrite indicates an increase in T in because of the high melting temperature of the compound.Nevertheless, the other phases in the concentrates also affect T in .
The exothermic heat Q ex values were obtained by the numerical integration of the exothermic peaks on the DTA curves (Figure 6).The exothermic heat is gen sociated with the oxygen content present during the test.However, in this case, the oxygen concentration is constant.Thus, it is possible that the higher value is due to the chalcopyrite (CuFeS 2 ) and coveline (CuS) contents.
The DTA and TG curves show that the sample-B reactions develop with low Q ex , which is associated with a signifi mperature and 1000˚C.A gradual weight loss in the range of 3 c particular, sample E exhibited a drastic loss of mass at 500˚C; this is because the range 450˚C -550˚C favours the direct oxidation of pyre to loadstone, while at temperatures above 650˚C the complete elimination of any formed ferrous sulphate is favoured; this is manifested as weight loss [16].
A gradual increase in mass up to 600˚C was observed for all of the conc e sulphides into oxides coincides with these masschange exothermic peaks (in DTA curves).A new and pronounced loss of mass up to 750˚C was observed for the five concentrates.At these temperatures, the trans-2 The proposed reaction mechanisms, based on the behaviour indicated by the DTA and TG curves and the equilibrium information obtained by calculation using the software progra able 4.
The DTA results are consistent with proposed reaction 5, which implies a loss of mass because of the sulphide-to-oxide transformation.In addition, it is important to consider 2 eactions 3 and 4).The initial mass loss revealed by TG is based on the decomposition of a pure Cu 2 S sample and the oxidation of pyrite, as observed in reactions 5 and 6.
The gradual increa st at 600˚C was also reported by Perez-Tello et al. [17].They believed that the concentrates quickly gain mass through the copper sulphate formation.Howev shows that reactions 11 and 10 contribute to the increase in mass through the formation of iron sulphide.Although this species is later dissociated, as indicated by reaction 14, the reaction speed is slow due to that the formation of sulphates, which closes the pores of the particle and prevents the reaction from occurring, and thus creates a gain in mass, as shown in Figure 7 [16].
The gradual final loss of mass is explained by the decomposition reactions of the sulphate into oxides.
Samples B and C showed the same behaviour during heating because of their similar mineralogical compos n: high content of chalcopyrite (CuFeS 2 ) and low copper and iron sulphides in the concentrate.Thes es developed intermediate T in and Q ex during heating.Sample A presented a lower T in and higher Q ex because of the higher contents of chalcopyrite (CuFeS 2 )  The microscopy study of the five concentrates solidified samp d iron sulphide (i.e., sample A) in the concentrates increase with the thickness of the white layer.In all samples, this white intermediate layer featured native copper bands or globular particles because of the decomposition of sulphides.In this case, the (Cu, Fe)S exhibited 31% lar crystals of (Cu, Fe)S in a matrix with a higher iron sulphide (FeS) content were identified.These phases were .(Cu, Fe)S contained approximately 52% Cu.Based on the S-Cu-Fe system, it was possible to conclude that different sulphide phases formed at the selected melting test conditions (1200˚C, 1 h) in both layers (A B or D B in the white intermediate layer an G of the grey top layer) contain up to 52% copper.Chalcopyrite decomposes at 623 K to an intermediate solid solution (ISS) (see Figure 10) or to chalcopyrite with a slightly different composition (CuFeS 2ssb ); moreover, an iron sulphide transformation occurred.The presence of native copper in all of the samples indicates that sulphur was only partly removed.

Conclusions
To summarise, the characterisation of the five concentrates studied in this work revealed th contained about 14 was the richest in were sulphides: chalcopyrite (CuFeS 2 ) and coveline (CuS).Iron was present in the concentrates in the form of FeS 2 or FeS.Samples A and E showed the highest iron sulphide contents The thermal analysis techniques applied in this study provided information that allowed us to determine the initial oxidation temperature T in and the exothermic heat Q ex of the samples, as well as the mass changes during he rogram HSC (Outokumpu), it was po e Minería," Secretaria de Minería de la Nación.Buenos Aires, Argentina, 1994, pp. 5 [2] Segemar, Serv e Argentina.
zan, "Algunos Fundamentos de las ción P. Sarquis and E. Brandaleze, "Factibilidad ating to 1000˚C.By correlating the DTA and TG results with the crystal phases identified by XRD and the information obtained though the thermodynamic equilibrium study performed using the software p ssible to identify the complex reaction mechanisms involved during concentrate heating.The reaction evolution features an initial transformation of the sulphides to oxides; then, the oxides react to form sulphates and are finally transformed to oxides.These results were confirmed by the phases identified by microscopy and the thermodynamic study on the Cu-Fe-S system.

Figure 6 .
Figure 6.DTA curves of the five copper concentrates samples.

Figure 8 .
Figure 8. Three layers present in the structure of solidified product A.

Figure 9 .
Figure 9. Copper band observed in the intermediate.Sample D contained the lowest concentration of Cu (14 s. oreover, the S/Cu ratio of sample D was 2.65, which is hi

re grey layer Table 3 .Figure 10 .
Figure 10.Sulphide composition determined in the white ( B ) intermediate layer and in the grey ( G ) top layer of sample

Figure 11 .
Figure 11.Mass evolution with respect to reaction mechanisms.n sulphide (FeS), promoting an intermediate hange in weight.bination of higher Q ex and T in .The m