Development of liquid–liquid extraction processes for concentrated hydrometallurgical solutions

Abstract

Hydrometallurgical liquid–liquid extraction is a separation and purification method used for the recovery of metals from aqueous solutions. In this thesis, process design methods were studied and evaluated for the development of liquid–liquid extraction processes for concentrated sulfate solutions. The conceptual process design was based on batch equilibrium experiments, theoretical equilibrium calculations, McCabe–Thiele analyses, pseudo-countercurrent extractions, and continuous experiments in industrial-type laboratory-scale mixer-settlers. Three different separation processes were studied: I) separation of arsenic from concentrated H2SO4, II) purification of a concentrated MnSO4 solution, and III) separation of cathode metals from lithium-ion battery (LIB) leachates.

Extraction of arsenic from concentrated (10.4 M) H2SO4 by undiluted tri-nbutyl phosphate (TBP) is accompanied by significant changes in phase densities and volumes, which were quantified. TBP extracted 83.7 % arsenic and 31.4 % H2SO4 from the H2SO4 solution in three countercurrent extraction stages.

Bis(2-ethylhexyl) hydrogen phosphate (D2EHPA) and bis(2,4,4- trimethylpentyl)phosphinic acid (BTMPPA) were used to purify concentrated MnSO4 leachate that contained over 150 g L−1 manganese. More than 99 % of the zinc and 98% of the iron were removed by both extractants in two countercurrent extraction stages. However, BTMPPA co-extracted less manganese than D2EHPA. Calcium was not extracted due to the high manganese-to-calcium ratio of the leachate.

A liquid–liquid extraction flowsheet was developed for producing nickel-rich sulfate mixtures from cobalt-rich LIB leachates. The nickel-rich solutions can be re-used in the synthesis of new cathode precursors. Additionally, purified Li2SO4, CoSO4, and MnSO4 are obtained simultaneously from the process. The benefit of the process is that no additional inputs of lithium, nickel, cobalt, or manganese are required. None of the developed processes are optimized. The future research could focus on process optimization through numerical flowsheet simulations based on mechanistic equilibrium models in which non-ideality and organic phase speciation are considered.