Palladium adsorption from alkaline glycine-cyanide solutions using activated carbon

There is an increasing demand for the use of Platinum Group Metals (PGMs) in various industries such as fuel cells, auto catalysts, electrical and electronics, e-waste and medical, which has significantly increased their market value (250 % increased value for palladium). Currently, technologies for separating and purifying metals, such as solvent extraction (SX) and ion exchange (IX), are generally applied with cyanide and acidic chloride systems to recover PGMs. In recent years, applications of amino acids, particularly glycine, have been developed and patented, as an alternative lixiviant for gold and copper leaching. Recent research on e-waste has shown an effective dissolution/leaching of Pd from glycine and glycine-cyanide solutions. The alkaline amino acid leaching represents an opportunity for a more selective and environmentally friendly approach, compared to the current applications using cyanide and chlorine systems.

The objective of this research is to investigate the feasibility of palladium (Pd) adsorption, using activated carbon, from synthetic leach solutions containing alkaline glycine–cyanide. 24-h tests were performed to assess the impact over Pd adsorption from various operational parameters including Pd initial concentration, pH, glycine and cyanide concentrations, carbon dosage, and temperature. Results showed that Pd can be efficiently recovered from glycine–cyanide solutions using activated carbon under different operational conditions, with the highest recoveries achieved at 20 g/L carbon dosage and pH 9.5 (98.2 %), and 95.5 % at 5000 mg/L glycine. The results have shown that the most significant parameters impacting adsorption efficiency (kinetics) are carbon dosage, temperature and pH. The experimental data from various initial concentrations of Pd were best fitted to the Langmuir model (R² > 0.9), with the maximum uptake capacity (Q_m) at different initial Pd concentrations ranging between 0.5 mg/g and 2.2 mg/g. In addition, the results of the kinetic modelling were best described by a pseudo-second-order model (R² > 0.99), for all parameters studied.