Evaluation of theMerrill–Crowe process for the simultaneous removal of platinum, palladium and gold from cyanide leach solutions

Abstract

Studies were conducted to evaluate the Merrill–Crowe cementation method for the simultaneous extraction of platinum, palladium and gold associated with copper and nickel from cyanide solutions, particularly high temperature eluates from carbon adsorption and desorption process. The paper reviews the Merrill–Crowe method and its potential for platinum and palladium recovery in the presence of gold, copper and nickel. In a mechanically stirred tank reactor and in the presence of pure nitrogen gas, statistically designed experiments were performed at two levels of the five operating parameters which were the free cyanide concentration, temperature, cementation time, base and precious metal concentrations and their studied ranges were 50–150 ppm, 60–90 °C, 15–90 min, 175–1350 ppm and 220–880 ppm, respectively. The optimum experimental conditions resulted in a solution with free cyanide concentration of 150 ppm at 60 °C and a reaction time of 90 min, base metal concentration of 175 ppm and precious metal concentration of 880 ppm. Under these conditions, cementation yield obtained was 91; 54; 19% for Au(I), Pd(II) and Pt(II) respectively. Zn addition was 1.5 times the stoichiometric amount for complete cementation for the low level case, while it was 0.68 times less the stoichiometric for the high level, in order to monitor competitive aspects of cementation from a mixed solution on a time scale and at conditions that might be relevant to industrial operationsIn contrast to Au cementation reaction, Pd and Pt cementation did not go to completion even though excess zinc powder (0.15 g Zn i.e. 1.5 to 3.4 stoichiometric ratio) was used with their mono-ionic solutions, which gave 48 and 36% of Pd(II) and Pt(II) respectively. In preliminary kinetic studies, deviations from the first-order rate law were observed. The rate of cementation of Au(I) was about 5 times higher than for Pd(II) and 15 times higher than for Pt(II), based on the apparent rate constant, k'1. It was also found that the ratio k'1/k'2 was smallest for Pt, in other words the deviation from 1st order linear kinetics was greatest.