Mr. Lars Felkl, Dipl.-Ing.

Abstract:

During secondary copper production both internal slags and a final copper slag is produced. In contrast to the final copper slag, which is a process by-product, internal slags are typically recycled to the preceding aggregate. Therefore also the part of copper and other valuable elements like nickel, tin, lead and zinc reporting to the final slag is kept low. On the other hand, other elements which influence the process negatively like aluminium and chromium are also recirculated and lead to an increase of density, melting temperature and therefore also viscosity of the resulting slags. This results in a more complicated slag treatment and influences aggregate capacity, i.e. (black copper smelter and the converter) and copper loss to slag, negatively.

This study is investigating a slag treatment (slag reduction) of high-copper bearing slags from secondary copper production to meet the described challenges by avoiding internal slag recirculation. But besides that, the process results to additional value in that the secondary slag produced through appropriate fluxing (i.e., through slag design) can be easily used for construction purposes. For this investigation an in-depth knowledge about not only the thermodynamics but also kinetics of this process is required.

This study is built on three pillars: Thermodynamical modelling, kinetic investigations and experimental tests in a medium scale. The thermodynamical modelling is conducted using FactSage™ with inclusion of the copper database. An open-system approach is used to model the reduction process by hydrogen. Through that several simulation steps are taken into account, where the hydrogen is added in fractions which are forming a thermochemical equilibrium with the slag and the resulting metal phase. After reaching equilibrium (during each step) the gas phase is removed and a new gas phase is formed by a new addition of hydrogen and the creation of a new thermodynamical equilibrium.

Kinetics investigations were performed by using thermogravimetric methods with the combined analysis of the off-gas stream using mass spectrometry for the achieved ratio of hydrogen to steam. This ratio can be used as a measure for the reduction progress. Firstly, the slag under investigation is mixed with hematite (Fe₂O₃) and silica (SiO₂) to achieve a secondary slag near fayalitic composition (45-50 wt.-% FeO and 35 wt.-% SiO₂). The non-fluxed slag would for be rich in alumina but contains also a non-negligible proportion of chromium(III)-oxide. Both compounds are known for increasing slag viscosity. The chemical composition of the resulting metal phase and the secondary slag is analysed using SEM-EDX.

The experimental trials in medium scale are performed with around 0.5-0.75 kg slag material, depending on the mass of needed fluxes. Here the gas is injected via a lance in the molten slag system at given temperature. Via weighing before and after the experiment the mass loss during reduction, i.e., due to zinc and lead fuming can be estimated. By means of chemical analysis of the achieved metal phase as well as of the resulting secondary slag the reduction degree in this scale can be evaluated. The main goal is to achieve a secondary slag which contains less than 1 wt.-% of copper and other valuable elements and a secondary slag with an iron oxide content and silica content of 45-50 wt.-% and 35 wt.-%, respectively.

It can be seen that the majority of slag reduction is completed within a few minutes and is therefore faster than when using carbon monoxide as a typical reducing agent, as long as diffusion can be neglected. In reality, this is not the case: due to the distance e.g., between the lance tip and the outer diameter of the reaction vessel, the necessary reduction time is extended, as the process is increasingly diffusion-controlled.

With increasing temperature, an accelerated reduction can be observed due to a reduced viscosity of the slag and an improved mobility of the hydrogen, whereby the reduction of the hematite added as an additive can be considered complete even before the reduction of the slag.