There are few methods to predict ion removal using ion flotation without lengthy experiments. The objective of this study was to model the Zn(II) flotation using a hybrid neural network/genetic algorithm (GANN) and multivariate linear regression (MLR). Mean square error and correlation coefficient values of 0.9228 and 190.1, respectively, for testing the datasets of the GANN model reveal the superiority of the GANN model in predicting the Zn(II) removal, while these values were obtained as 0.9125 and 220.36, respectively, for the MLR model. The results showed that this model could be useful in predicting the Zn(II) removal in response to changes in the input parameters.

The present study deals with nickel ions removal from dilute aqueous solution by ion flotation with emphasizing the process efficiency. The effect of collector structure on ion flotation efficiency and water recovery was evaluated using anionic collectors of sodium dodecyl sulfate (SDS) and...

Millions of tons of zinc plant leach residues (ZPLR) have been stockpiled in Iranian hydrometallurgical zinc plants during the last few decades. Due to the low grades of zinc, lead, and silver in these residues, these residues have been abandoned without treatment. The authors of this paper studied zinc plant leach residues (ZPLR) to propose a flotation process for separating and producing lead and silver concentrate. A response surface methodology (RSM) was employed to obtain six models for optimizing the best conditions for lead recovery, lead grade, zinc recovery, zinc grade, silver recovery, and silver grade. In these models, the effect of the different main variables, including density, flotation time, pH, sodium sulfide dosage, and potassium amyl xanthate dosage, was investigated to optimize grades and recoveries. The studied ZPLRs were categorized into two types based on the disposal time, including new and old residues. The chemical analysis showed that the grades of lead, zinc, and silver in the new residues are higher than in the old residues. In a previous mineralogical study, it was found that silver forms in lead and zinc minerals as a solid solution within their structures. The resulting 3D graphs showed that the interacting variables have significant effects on responses. The ANOVA analysis exhibited the order of model significance to be lead grade (F-value of 36.46) > silver grade (19.76) > lead recovery (7.88) > zinc grade (5.63) > silver recovery (5.58) > zinc recovery (4.83). Based on these models, under the conditions of 1126.26 g/cm³ density, 20.83 min retention time, 9.9 pH, 6 kg/t sodium sulfide, and 749.66 g/t potassium amyl xanthate dosage for a new residue type, the recoveries of lead, zinc, and silver were determined to be 51.10%, 11.13%, and 72.85%, with grades of 38.87% Pb, 8.46% Zn, and 1209.11 g/t Ag, respectively. According to the feasibility study results, the presented work is reasonable in terms of technical, economic, and investment potential.