Purification of VOSO4.5H2O using solvent extraction

Abstract

Vanadium (V), which is found in over 50 minerals usually in combination with other elements is used in a variety of applications including cladding for titanium to steel, material in nuclear fusion reactors, ceramics, glass coatings, catalysts, automobiles, rubbers, plastics, chemicals, spacecraft and airplane industries and submarine pipes. Apart from these existing applications, a novel application of V is in the all vanadium redox flow battery, which has shown promise as a future energy storage system. All these applications have given rise to an increased V demand and the subsequent opportunity for the development of new or improved purification methods. In view thereof, the aim of this M.Sc. study was to develop an improved solvent extraction (SX) based purification method for the recovery of VOSO4.5H2O (V(IV)) from a multicomponent electrolyte solution containing Fe2(SO4)3, Al2(SO4)3, Co2(SO4)3.7H2O, NiSO4.6H2O and K2CrrO7. After optimisation of the SX variables, the suitability of membrane-based solvent extraction (MBSX) was investigated. The research presented can be divided into three parts. In a first step two preliminary experiments were conducted, where i) the extraction of V from different H2SO4 feed concentrations using eight extractants and ii) the effect of the impurities in the synthetic solution on V(IV) extraction was investigated. In the second step, the SX process was optimised in terms of i) the extractant concentration, ii) the effect of contact time, iii) the effect of diluent, iv) the effect of organic and aqueous volume phase ratios, v) the effect of acid concentrations, vi) the effect of the stripping agents and vii) stripping agent concentrations. In the third part, MBSX experiments were done using the optimised SX conditions. From the preliminary experiments, it became apparent that of all of the eight extractants that were tested, the oxime based extractant, LIX-84-IC and the tertiary amine based extractant, Alamine 336 gave the highest extraction of V(IV), specifically at low acid concentrations (0.1 mol/dm3). The results from the single salt extractions confirmed the complexity of the V-speciation in aqueous solutions as well as the influence of the impurities on extraction. Since both LIX-84-IC and Alamine 336 showed promising results, the SX process was optimized for both the extractants as discussed in Chapter 3 and 4 respectively. For LIX-84-IC, the highest V extraction was attained when extracting for 15 minutes using a 30 % (v/v) LIX-84-IC containing organic phase consisting of cyclohexane and 1-octanol as the diluent and modifier. For this extraction the O/A volume ratio was 1:1. While under these conditions the co-extraction of Cr(VI) was observed, little to no-co-extraction of the other salts occurred. The selective stripping of V(IV), without any co-extraction fo Cr(VI), was achieved using a 0.05 mol/dm3 NaOH solution. When using MBSX significantly higher equilibrium times were required while the selectivity decreased in terms of the V(IV) extracted. The delayed time suggests mass transfer limitations when using LIX-84-IC. When using Alamine 336, optimal extraction was obtained from low H2SO4 feed concentrations (0.0002 mol/dm3), confirming a metal complex anion exchange extraction mechanism due to the presence of metal anions in the feed solution. Optimal extraction was achieved when using 20 % (v/v) Alamine 336 with a contact time of 90 minutes using ShellSol 2325 and 1-octanol as the diluent and modifier and an O/A volume phase ratio of 1:1. While the co-extraction of Cr(VI) was again observed, selective V(IV) recovery from the loaded organic was achieved when using a 0.03 mol/dm3 HCl solution. Additionally, scrubbing with 7 mol/dm3 HCl can be used to first remove Cr(VI) from the loaded organic prior to the stripping of the V(IV). When using MBSX, both the selectivity for V(IV) as well the recovery increased compared to SX, which is contrary to the results obtained when using the oxime-based extractant, while no significant change in the equilibrium time was observed. The results for the batch extraction experiments showed that, when using the oxime based extractant LIX-84-IC, the purity of V(IV) increased from 15.02 % to 99.98 % in a single extraction and stripping stage with a yield of 88 %. When using the amine based extractant Alamine 336, the V(IV) purity increased from 15.26 % to 95.3 % with a significantly lower yield of 27.1 %, also in a single extraction/scrubbing and stripping stage. Hence while high purities of V(IV) were attained using both the extractants when using SX, the higher yield attained favours the oxime based extraction. Contrariwise, in the MBSX studies, Alamine 336 outperformed LIX-84-IC in terms of recovery and extraction time.