Abstract
This study presents a process that can separate cobalt and nickel from Base Metal Refinery (BMR) secondary leach liquor to obtain pure cobalt and nickel aqueous streams. The two metals are becoming more important due to their increased use in recent renewable energy and automation technologies. Their efficient recovery from process streams is therefore of great importance. Previous studies focused on the recovery of the two metals from primary leach solutions obtained from high pressure acid leaching of laterite ores, waste rechargeable batteries, scrap metal, and from synthetic solutions. The current study investigated the recovery of cobalt and nickel from BMR secondary leach liquor which is a weakly alkaline solution produced after the production of nickel powder in BMR flowsheets. This feed solution contained cobalt, nickel, ammonium sulphate and ammonia. The extractant used was (bis(2,4,4-trimethylpentyl)phosphinic acid (Cyanex 272)) diluted in kerosene.
In the first part of the study the factors considered were; Cyanex 272 concentration (5 – 25 v/v%), organic to aqueous (O/A) loadings (0.33 – 3), and extraction temperature (20 – 70 oC). The investigation was conducted to determine the optimum extraction conditions, optimum stripping conditions, the extraction mechanism and process thermodynamic parameters. The aqueous feed solution analysed; 404.49 mg/L cobalt, 1535.6 mg/L nickel, and 580 g/L ammonium sulphate (pH 7.2). The initial aqueous phase pH was fixed at 6.77 to avoid the precipitation of double salts whilst maintaining high cobalt selectivity. The three factors mentioned above were studied and optimized using the design of experiments (DoE) technique. The maximum separation factor and percentage cobalt extracted were 169.78 and 88.85 % respectively. A second stage batch extraction resulted in the complete isolation of nickel in the raffinate (985.14 mg/L Ni) whilst, selective stripping of the loaded organic phase resulted in a complete isolation of cobalt (1500.21 mg/L Co). It was therefore possible to separate cobalt and nickel from BMR secondary leach liquor under the following optimum conditions; 5 v/v% Cyanex 272 concentration, 0.5 O/A loading, and 60 oC extraction temperature. Fourier Transform Infrared Spectroscopy (FTIR) and Ultraviolet–Visible Spectroscopy (UV-VIS) analysis confirmed the formation of the tetrahedral cobalt-Cyanex 272 complex, the absence of the nickel-Cyanex complex and possible regeneration of the extractant.
ii
The study then investigated the use of ammonium-saponified Cyanex 272 on the separation of cobalt and nickel from the BMR secondary leach. This second part of the study was done to determine if partial saponification of the extractant improves the cobalt extraction percentage and the separation factor. The factors considered in this part were saponification percentage (5 – 25%), O/A loading (0.33 – 3), and extraction temperature (30 – 70 oC). These factors were optimized using the DoE technique and the maximum cobalt extraction percentage and separation factor were 94.07 %, and 1189.76 respectively. The optimum conditions for this new process were; 20% saponification, 0.5 O/A loading, and 60 oC extraction temperature. It was therefore concluded that partial saponification improved the extraction efficiency by enhancing cobalt extraction, and limiting nickel co-extraction. This was largely possible due to the maintenance of equilibrium pH at the optimum saponification percentage. It was also concluded that saponification does not affect other factors such as O/A loading and optimum extraction temperature. However, the cobalt and nickel hexamine complexes were more stable in the saponified system as extraction was significant above 50 oC unlike in the un-saponified system where there was significant extraction below 50 oC.
Lastly, a mechanistic mathematical model based on the stabilities of complexes in both the aqueous and organic phases, and the activities of species in the aqueous phase was developed. The model was solved in Matlab and results were fitted to experimental data using the curve fitting tool. Non-linear regression through the curve fitting tool enabled accurate estimations of the unknown parameters and the model produced good fit statistics. It correctly predicted the influence of initial extractant concentration on cobalt extraction percentage, which is novel since previous models only work at a certain specific extractant concentration. The model also correctly predicted the influence of O/A loading on both cobalt and nickel extraction.
In conclusion, the new process conditions presented in this study enabled successful separation of cobalt and nickel from BMR secondary leach liquor, or any similar aqueous solutions. The mechanistic model presented in this study is also new and it reasonably explains the selective extraction process well. Future studies should focus on speciation in the initial aqueous phase at different pH to gain more understanding of this feed solution and improve the extraction efficiency. Studies should also focus on generating equilibrium data using aqueous solutions that are similar
iii
to the BMR secondary leach liquor considered herein. Other future studies should focus on process design and simulation as well as pilot plant scale investigations.