Abstract:
A novel method has been developed whereby certain high phosphorus, manganese ores of eastern coastal region of India can be dephosphorized by roasting the ores with small quantity of sodium chloride followed by leaching in dilute solution of mineral acids. In this paper an attempt has been made to study the leaching behavior of roasted ores from the point of view of statistical design of experiments. The results indicate that the three main variables, namely, concentration of acid, pulp density and the temperature of leaching and their mutual interactions significantly influence the extent of dephosphorization. The linear regression has been found to describe the leaching data better than the method of analysis of variance. Statistical method of optimization appears to be difficult and therefore random search method has been used to arrive at the optimum conditions of leaching.