Molecular modelling of species pertaining to the solvent extraction of tantalum penta-fluorides

Abstract

Solvent extraction (SX) is used for the separation and purification of various metals, including tantalum (Ta) and niobium (Nb). Industrial processes for the separation of Ta and Nb traditionally use high concentrations of hydrofluoric acid (HF), sulphuric acid (H2SO4) and extractants including methyl isobutyl ketone (MIBK), making this process dangerous and detrimental to the environment. Ungerer et al. studied the separation of Ta and Nb, investigating safer chemicals and alternative techniques. During this study, separation was achieved in a H2SO4 medium using the extractants diisooctyl phosphinic acid (DioPA) and di-(2-ethylhexyl) phosphoric acid (D2EHPA). The main obstacle during this study remained the speciation of Ta and Nb, springing the question of why separation occurred with some extractants and not with the others. One method for determining the speciation of a reaction is by using computational techniques for molecular modelling. Progress in computational chemistry over the last 20 years has made quantum mechanical calculations on large molecules, chemical systems as well as on macromolecule reactions possible. Calculations based on the density-functional theory (DFT) are now not only used on light elements and small molecules, but also on metal complexes, heavy metals and especially on metal separation in solvent extraction. The main current goal of computational methods for SX is the analysis of the extraction process on the molecular level, determining the molecular reactions as well as the system reactions occurring during SX from a thermodynamic point of view and thereby developing new methods for whole system analysis of the SX process of metals. The advances in computational chemistry consequently provide the possibility to determine with good approximation the outcome of proposed SX experiments before embarking on expensive, time consuming experiments and environmentally harmful waste generation. To investigate the suitability of modelling for this application, a case study (Part 1) was selected where it was hypothesised that when TaF5 is dissolved in water, it could react stepwise with water to finally form tantalum penta-hydroxide (Ta(OH)5) and other oxyfluoride species including TaOF3. Due to the fact that literature on TaF5 reactions with water is limited, TaCl5 and its reactions were used to develop the model (method). As part of the model development and verification, DFT was used to calculate the energy needed for these reactions, comparing different functionals and basis sets. The validated model was then applied to TaF5 as a case study. From the results it was confirmed that the reaction of TaX5 (X = Cl or F) with water to form Ta(OH)5 and Ta2O5 is an endothermic reaction, while the formation of Ta(H2O)F5 and TaF4OH was exothermic. The next step (Part 2) in the study of the aqueous phase was to calculate the energy needed for various reactions of H2SO4 and H2O in an aqueous phase. Again different functionals and basis set combinations were used and compared. According to the results, the deprotonation of H2SO4 was endothermic in a 1:1 acid-water ratio, exothermic forming HSO4- in a 1:5 acid-water ratio, while SO42- formed exothermically by a double deprotonation in a 1:10 acid-water ratio. Furthermore, it was seen that hydration and dehydration of H2SO4 in a bulk H2O solution was a continuous process. From the energy calculations it was determined that although the H2SO4.H2O, HSO4-.H2O and H2SO4.2H2O species could form, they would most likely react with H2O molecules to form HSO4-, H3O+ and H2O. The next step (Part 3) combined Part 1 (TaF5 + H2O) and Part 2 (H2SO4 + H2O). The results obtained were used to attempt to predict the reaction mechanism occurring during SX. From previous modelling it was seen that by increasing the number of water molecules, the reaction energy decreased due to molecule stabilisation (hydrogen bonding) and subsequently a 1:1:10 metal:acid:water ratio was used. Results showed that in a 1:1:10 metal:acid:water ratio the deprotonation of H2SO4 was exothermic, leading to the formation of HSO4- and a lowering of the reaction energies from being endothermic to between -40 to -103 kcal/mol. Furthermore, from the various reactions and geometries between TaF5, H2SO4 and H2O investigated, it was observed that only three species will be available in the aqueous phase during solvent extraction, namely TaF5.H2O in a water or diluted acid medium, TaF4.HSO4 in a concentrated H2SO4 medium and TaF3OH.HSO4 if the aqueous phase aged. In an attempt to understand how extraction occurs, molecular dynamic simulations were used, whereby each species (identified in Part 3) was simulated in a 3D periodic box. The stoichiometry of each system was determined from previous experimental (SX) conditions and each species was investigated at 4 and 10 M H2SO4. Simulations started at a perfectly mixed point. The small-scale system results showed that TaF5.H2O forms at low H2SO4 concentrations and can be extracted with D2EHPA in both 4 and 10 M acidic conditions. The ageing of the aqueous phase leads to the formation of TaF4OH, which cannot be extracted with D2EHPA at either concentration. An H2SO4 medium leads to the formation of TaF4.HSO4, which could be extracted with D2EHPA from both 4 and 10 M H2SO4. The ageing of this solution results in the formation of TaF3OH.HSO4, which could not be extracted at either H2SO4 concentration. Furthermore, it was seen that, in the 4 M H2SO4 system, the aqueous phase tends to form a droplet within an organic bulk solution and when the H2SO4 concentration increased, both phases showed droplet properties with break-aways between the phases.