Exploration of the thiosulphate process for the dissolution of gold from electronic waste and its recovery through ion-exchange

Master Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title
Electrical and electronic equipment (EEE) has substantially grown over the past few years due to vast technological advancements and consequently so has electronic waste (e-waste). This growth has shown cause for concern with a generation of 53.6 million metric tonnes of e-waste globally in 2019 alone. Part of this concern may be due to the slow adoption of formal collection and recycling practices in developed countries whilst developing countries bear the brunt of the e-waste produced within the country and the e-waste imported from other countries. Furthermore, the burden of the increasing levels of e-waste has detrimental effects on the environment. In such cases, e-waste landfills have been known to leach metals, such as lead, into soil and groundwater of nearby regions thereby affecting human and animal life in the area. Despite the evident hazardous materials associated with e-waste, there is still value in this waste. The value is attributed to the metals such as gold, silver, copper and palladium found in e-waste. Printed circuit boards (PCBs) are a small but nonetheless crucial fraction of global e-waste making up 6% by mass. The gold content in this small fraction is of much higher concentrations than in typical primary gold ores thus PCBs represent the most economically attractive portion of e-waste. Hydrometallurgical processing has previously been applied for the recovery of gold and copper from primary ores possibly due to it being considered environmentally friendly. Therefore, this thesis investigates a hydrometallurgical process for recovering gold from PCBs. Literature studies show ammonium thiosulphate to be a viable option in comparison to the more widely used cyanide leaching route (Aylmore & Staunton, 2014). Thus, the ammonium-thiosulphate system containing ammonia, ammonium thiosulphate and copper (II) sulphate pentahydrate was incorporated as a hydrometallurgical option for the recovery of gold. This study explores the formulation of synthetic gold solutions using gold powder as well as its application to PCB gold leaching in the same ammonium-thiosulphate system. Gold recovery using ion-exchange processes was investigated with the use of a medium-base (AuRIX®100) and two strong-base anion exchange resins (Purogold™ MTA5013SO4 and MTA5011SO4). The AuRIX®100 resin was specifically developed for and is currently used in the gold-cyanide system. The MTA5011SO4 resin is currently used in the Barrick Goldstrike operation for thiosulphate systems and the MTA5013SO4 resin was produced by Purolite® for thiosulphate-copper systems. In addition, two eluants (ammonium nitrate and ammonia) were tested to effectively elute gold from the resin. Furthermore, copper was monitored throughout the leaching and ion exchange experimentation due to its catalytic effect in the ammonium-thiosulphate system. Determining the most appropriate dilution of the ammonium-thiosulphate solution after leaching to ensure solution stability in the periods between sampling time and analysis was explored as part of the investigation. In addition, gold powder dissolution from a 99.99% pure gold powder and gold leaching from PCBs were investigated. PCBs were cut by means of a bandsaw for the purposes of fitting them into a reactor whilst limiting copper liberation. Various additional background copper concentrations were introduced into the gold powder dissolution and leaching system to determine its effect on gold and copper extractions. For the ion exchange processes, capacity tests on all three resins were conducted to establish the resin operating capacities before loading and elution of synthetic gold solutions. Loading and elution tests were conducted at three flowrates (10 mL/min, 25 mL/min and 50 mL/min) and with two eluants; ammonium nitrate and ammonia. Kinetic and equilibrium experimental work was investigated on the MTA5013 resin with the addition of chloride ions as a competing anion to loaded gold on the active sites of the resin. The MTA5011 resin was introduced into the experimental work for confirmatory results of the MTA5013 resin. This was necessary as both the Purogold™ resins are similar with the exception of particle size and thus were expected to behave in the same manner. Gold concentrations of 141 ppm (representing 100% extraction) and 4.1 ppm (representing 91% extraction) after 24 hours was extracted for gold powder dissolution and PCB leaching experiments respectively. Background copper concentrations of 0.045 M and 0.1 M resulted in the highest gold and copper extraction values for gold powder and PCB respectively. In addition, the strong-base anion exchange resin (MTA5013) proved to be more suitable than the medium-base anion exchange resin for the ammonium-thiosulphate system. However, its low gold loading values of 8.06 meq/L (milliequivalents per litre of resin) and elution of 5.80 meq/L proved that it was ineffective in removing large gold amounts from the synthetic solution at a resin volume of 5 mL. Loading at low flowrates of 10 mL/min whilst eluting at 25 mL/min resulted in the highest loading and elution concentrations of both gold and copper. Copper loading and elution concentrations of 150 meq/L and 38.4 meq/L respectively were measured. This translated to 3.78% recovered in loading and 67.5% recovered in elution. High recoveries of gold and copper representing 75% and 67.5% respectively were achieved with ammonium nitrate as an eluant. Only 21.7% of the loaded gold was eluted when using ammonia as an eluant. Final kinetic and equilibrium test work revealed that the MTA5013 resin has an affinity for the aurothiosulphate ion over the weak chloride ion. This is attributed to low concentrations of chloride in solution therefore being ineffective in displacing the aurothiosulphate ion from the resin. Results from this study suggested that anions such as thiosulphate and tetrathionate competed strongly for the resin active sites and this was in agreement with Nicol & O'Malley (2002) who made the same postulation. Moreover, final gold loading concentrations were low (8.06 meq/L) given the operating capacities of the resin (0.77 eq/L). It was proposed that this may be due to other anions such as polythionates in the system occupying the active sites including the small resin volume. High concentrations of polythionates in solution compete for resin active sites despite the resin affinity for the aurothiosulphate ion over polythionates in the system. It is concluded from this study that the ammonium-thiosulphate was efficient in leaching gold, obtaining almost 100% extraction in PCB leached solutions and 100% in gold powder synthetic solutions. Strong-base anion resins were proven to be ineffective in obtaining high gold recoveries for the system at low resin volumes and the resin indicating a high concentration of polythionate attachment relative to gold. However, the resin did demonstrate a higher affinity for the aurothiosulphate ion relative to copper species in the solution and given a larger resin volume: solution volume ratio, high gold recoveries are possible. Furthermore, an ammonium nitrate eluant is considered appropriate in removing high concentrations of gold from the resin.