Matte, oxygen content

The presence of oxygen in the system leads to the formation of a Fe-S-O ternary matte because of the miscibility of FeO with FeS (2,3). The oxygen content in the Fe-S-O matte dqiends on the oxygen potential or CO/CO2 ratio of the system. 

Test-7 was done at CO/CO2M. The average matte composition is Fe/S=3.77 and 6.51% O. The zinc solubility is lower than that in Test-6 because of the lower oxygen content of the matte or the lower oxygen potential. 

Figure 9 summarizes the influence of the oxygen content on the zinc solubility for Tests-5, 6 and 7 having a Fe/S ratio between 3.3-3.7. It is seen that the solubility of zinc is directly proportional to the oxygen content of the matte. 

From the oxygen contents reported in Table IV for Tests 6 and 7, an excellent agreement between the thermodynamically predicted oxygen contents of the matte was foimd. This was obtained from the initial matte composition (Table IV), and the measured data, shown as the average matte composition for different CO/CO2 ratios. Consequently, the theoretical analysis was successfully verified. For example, at C0/C02=2, Fe/S=3.3 and 1350°C, thermodynamics predicts 9 wt% O and azns=0.05 in comparison with the measured 8.61 wt% O and 0.49 wt% Zn in Test-6. Furthermore, the experimental data confirmed that the activity coefficient of ZnS(l) in the matte is about 10. 

Table V summarizes the experimental results of the tests with low oxygen content in the matte i.e., tests 1-5,9and 10. The matte conqxrsititMi and nramalized zinc solubility at Pzn=0.1 atm and 1350°C as reported Table III are repeated here. The zinc solubility has been converted into mole fractions to calculate the activity coefficient. The sulfur potentials at 1450 and 1200°C are taken from the reported activity diagrams for Fe-S-O matte (9), from which the sulfur potentials at 1350°C are estimated. The activities of ZnS in the matte were calculated from azns = K Pzn (Ps2), where K is the equilibrium constant for reaction Zn(g)+V2S2(g)=ZnS(l) and was estimated to be 64 at 1350 C. The activity coefficient of ZnS in the matte was estimated (Table IV). Although the estimated activity coefficient of ZnS varied from 7 to 22, it has an average value of 12.6 and is very close to the reported value, 10 (8). Considering the many possible sources of aror in the tests, it is not surprising that there is a large scatta in the estimated ZnS activity coefficient Nevertheless, the experimental data wae in close agreement with the theoretical predictions.

The melt copper (matte) is further processed in a converter, where the remaining sulfur and iron is removed by underbath injection of oxygen-enriched air. The resulting so-called blister copper has a copper content of 98%. 

---