Optimization of the treatment process of zinc leaching residue by using the response surface method

2.1 Materials

According to previous X-ray powder diffraction (XRD) analyses, the leach residue used in the experiments contains 18% zinc, 6.3% SiO₂, and 1.76% CaO. The zinc in the residue is mainly in the form of zinc ferrate (ZnO·Fe₂O₃) and a little bit of zinc sulfate (ZnSO₄). The CaO added as a slagging agent was analytical pure. The fixed carbon content of the reduction carbon powder was 78%.

2.2 Experimental procedure

The zinc leaching residue was dried at 100°C and mixed proportionally with CaO and carbon powder. The reduction of the volatilization process of zinc was simulated in a ceramic crucible with a cover at 1200°C for a certain time; at the end of the experiment, the crucible was cooled to room temperature in the furnace. Next, the residue was taken out and weighed, after which the zinc content was analyzed by using the EDTA volumetric method (according to the China National Standard GB/T 14353.3-2010). This was done to calculate the zinc volatilization rate.

2.3 Experimental design and results

The response surface experiment design was carried out using the response surface design platform of JMP (USA), a data statistics software. Three parameters were considered (carbon content, basicity, and holding time) and the setting of each parameter is shown in Table 1. The variable values of nos. 1 to 16 were designed by using the central composite design (CCD) and nos. 17 to 24 belonged to the extended experiments. The experimental results are shown in Table 2.

3.1 Statistical analysis

The stepwise regression method was chosen to analyze the experimental data in Table 2, and the quadratic response surface model was established after eliminating the non-significant items. The model for zinc volatilization rate is calculated by the obtained quadratic regression equation given by

where X₁, X₂, and X₃ are equal to carbon content (wt.%)−3515,basicity−1.641.36, and holding time (min)−4515, respectively.

According to the analysis of variance and test for lack of fit (Table 3), this equation is suitable for expressing the model. The R²=0.91 and the high F value (35.0302) and low p-value (p<0.0001) indicate that the fitting model is reasonable. Hence, the model can be used to analyze and optimize the process of zinc leaching in the rotary kiln. The p-value (p>0.05) of the test for lack of fit also shows that the fitting model does not exhibit lack of fit. Table 4 shows the significance of each coefficient in the equation, which is determined by the p-value listed. As can be seen, the effects of the first term X₁, X₂ and the quadratic term X₁², X₂² on the response of zinc volatilization rate are significant. This means that the carbon content and the basicity are the main parameters affecting the rate of zinc volatilization, and do not indicate a simple linear relationship.

3.2 Reduction mechanism

Although the zinc ferrite in the zinc leaching slag cannot easily undergo thermal decomposition, they tend to be easily decomposed when added carbon or under the conditions of carbon monoxide atmosphere [11], [12], [13]. This can be explained by the following reactions expressed by Equations (2) and (3). Another zinc–bearing phase in the slag can also be reduced by carbon as shown in Equations (4) and (5).

The values of the Gibbs free energies in Equations (2) to (5) can be respectively estimated by using Equations (6) to (9). The values at the temperature over 1060.85°C are less than zero, indicating that the reduction reaction with carbon can be carried out spontaneously with a temperature higher than 1060.85°C. Thus, the effect of carbon will not be significantly larger when the temperature reaches this condition.

3.3 Optimization

The above response surface quadratic model was applied to explore the best combination of the three variables. The changes of zinc volatilization rate with different selection of variables are shown in the prediction profiler (Figure 1). As can be seen, the rate tends to be high with the increase of content of carbon. Until the content of carbon reaches 32%, the rate change is very small. While basicity has a more significant effect on the zinc volatilization rate, there exists a nearly linear relationship between them when basicity is higher than 1.5. Moreover, the holding time had little effect on zinc volatilization. Thus, it can be concluded that the zinc volatilization rate could achieve almost 100% with carbon content of 32% carbon, basicity of 3, and holding time of 30 min.

Figure 1:

Prediction profiler for the zinc volatilization rate.