Flotation Enhancement of sedimentary phosphate ores by cornstarch as an environmental depressant: Modeling and analysis using full factorial design (FFD) and artificial neural network (ANN) approaches

Abstract

The present work attempts to highlight different facets of corn starch as an environmentally friendly depressant using soybean oil as a collector. The important influence of parameters such as pH, collector and depressant dosage, and the role of the depressant are discussed. The grade (%P2O5) of the flotation products were analyzed by means of UV–visible spectroscopy, the recovery (%Re) and the efficiency (%E) of the flotation products are calculated based on the grade. The mechanism of action of starch depression was revealed through inductive coupled plasma (ICP), Fourier transform infrared (FT-IR) spectroscopy, and X-ray diffraction (XRD). In addition, full factorial design (FFD) and artificial neural network (ANN) were used to generate an evaluation approach for P2O5 content. Moreover, the results obtained confirm that starch has an influence on phosphate depression at more acidic and alkaline pH. Indeed, at pH ​= ​4, a P2O5 content of 28.29% was obtained with a recovery of 87.46% in the non-floating fraction. Similarly, at pH ​= ​12, a content of 27.60% P2O5 with a recovery of 92.10% was found at a CaO/P2O5 ratio equal to 1.6. These concentrates were obtained from a feed sample containing 22.09% P2O5 using 10.3 ​g/L of soybean oil and 15 ​g/L of corn starch. The results of the comparison between the ANN and FFD approaches show that the ANN model outperforms the FFD model in terms of performance, with a good and higher coefficient of determination (R2 ​= ​0.999).