Book Description

Copper is widely known as a very important material due to its applications in our daily life, such as electrical devices and heating appliances. It is not so common knowledge that copper is not found in its metallic form, but mixed with other metals and elements like sulphur and oxygen. The process to obtain pure copper nowadays implies a strong impact on the environment. Regarding copper sulphides, its reduction to metallic copper is based in the oxidation of the ore products which forms sulphur dioxide (SO2), amongst others. Although SO2 is sent to sulphuric acid production, there are still emissions to the environment. Since the reduction of contaminant emissions has become of a primary concern, several alternatives have been studied to replace the current process. One of the alternatives that is gaining strength is the leaching or bioleaching, where poor ores are treated with solvents (chemical leaching) or with bacteria (bioleaching). The advantages of this process is that low pressures and temperatures are needed, but long times are required. However, a high pressure and temperature version exists to reduce the leaching retention time. Another option is the so – called mechanochemical reactions. These are based in applying high energy by grinding and milling a small amount of the sample. The main disadvantages are that the reaction is slow and small amounts can be treated at a time. This choice is taken where tiny particle size is required, since copper of nano size can be prepared. In addition, it is used as a pre – treatment for the leaching/bioleaching process, because the milling activates the reactivity of the solid and shortens the leaching times. The process that has attracted more attention is the reduction with hydrogen (H2) gas. If feasible, a similar system as it exists nowadays could be adapted to the H2 flow. Regarding the particular case of copper sulphides, several authors have already investigated the reaction. Among their conclusions, the reaction is not thermodynamically favourable unless another material is added to capture the hydrogen sulphide (H2S) produced in the reducing reaction. Numerous studies have based the good conversion results of this reaction by the addition of a considerable amount of another substance, practically in all cases, lime, i.e. calcium oxide. The role of this compound is to react with the H2S produced in the reduction, forcing the reaction to the production of metallic copper. The purpose of this thesis was to provide more information about this particular case, the H2 reduction, because it seems to generate more interest due to its good results. This process can apply to both copper oxides and sulphides; however, this project was focused only in the reduction of copper (I) sulphide (Cu2S), commonly known as chalcocite. In addition, a new attempt to improve the bare reaction is made by mixing metallic copper on one set of experiments and copper (II) oxide (CuO) in another. Moreover, the modelling of the process was approached basing the mathematical solution on the development made by H. Y. Sohn for reactions between porous solids and gases. The materials used to perform the experiments were Cu2S, metallic copper and CuO, and the reactant gas was a mixture of 4% H2 and 96% argon (Ar). The rate of conversion was measured by weight loss in a conventional thermo gravimetric analysis (TGA) set up for the bare reduction reaction. In this particular set up, experiments with loose powder at different reaction temperatures were carried out, as well as experiments in the pellet form and mixtures with copper. These last cases were performed only at a concrete isotherm. A differential scanning calorimetric (DSC) set up was used to perform the same experiments as in the TGA set up, but in pellet shape, to confirm that no oxidation was affecting the results. The experiments mixing CuO were only completed in this set up. The results from the loose powder experiments gave low sulphur removal conversions, at a low partial pressure of reactant gas (H2 partial pressure is 0.04 MPa). The isotherm values showed different mechanisms for the high (750 – 850oC) and the low (600 – 700oC) temperature range studies. An attempt to determinate the kinetics of the chemical reaction was done taking the very first values of the reaction, where diffusion was not playing an important role. The findings were that the kinetics followed the shrinkage core model in the whole range, while the activation energy for the reaction was 61 kJ/mol for the low range and 8 kJ/mol for the high temperature range. Diffusion became strong quite fast after the chemical reaction stage. Its effect was studied by means of performing the exact same experiment in pellet shape. Since the porosity of the pellet is lower, diffusion was expected to be harder and lead to a lower value, but instead the sulphur removal from became higher. The action of compressing the powder introduced defects into the powder, making it more active which could explain this discrepancy. A step forward to previous experiments in the field was to add a certain amount of metallic copper with the idea of set nucleation spots for the copper production. All mixture conversions turned out to be lower than that for the simple Cu2S reduction, which might be due to a thicker product layer and faster shrinkage of the copper formed, which prevents the reactant gas to go further. Another approach was to produce the copper spots for nucleation in situ, with the reduction of CuO by H2 before the main reduction. In the test conditions, the experiment with lower CuO content (1 wt %) generated noticeably better overall conversion results than the normal reduction. On the contrary, mixtures with a higher content (e.g. 10 CuO wt %), in the test conditions, were not able to reduce the CuO completely before heading to the isothermal conditions of the main reduction, which lead to the production of SO2. As a result, the improvement in the conversion cannot be taken into account. An important part of the project is focused on the application of the model developed by H. Y. Sohn et al to describe the experimental results obtained from the isotherms. A first attempt considering only chemical reaction and diffusion effects fairly represented the conversion values in the intermediate region. The first stage is clearly controlled by chemical reaction, while flat slope of the latest period indicated the effects of mass transfer. The second attempt was to add the resistance of the gas film surrounding the pellet to the model. The new conversion curves were closer to the slope at the end of the experimental results, but differed deeply from the inital values. As main conclusions, it was seen that the temperature affected extremely to the mechanism of the reaction. The shrinkage core model described fairly well the chemical reaction controlled part, while the addition of diffusion or mass transfer gives better approximations to the latest stages of the process. The addition of metallic copper turned out to be badly for the sulphur removal because the product sinters faster, while the addition of CuO seems to improve the reaction, although more experiments should be done regarding the latter.