Effect of alkyl chains configurations of tertiary amines on uranium extraction and phase stability – part II: Curvature free energy controlling the ion transfer

Highlights

• The third phase quenching observed for longer or branched tertiary amines is associated to reduced attractive interactions.

• Thermodynamic analysis shows that longer or branched alkyl chains induce increased aggregates curvature free energy.

• Curvature energy becomes a predominant inhibitor of metal free energy of transfer.

• Metal transfer is decreased because of high entropic contribution originated by the extractants branching and packing.

Abstract

We consider a structural and thermodynamic analysis of uranium solvent extraction by tertiary amines for which the alkyl chain configuration has been modified. The first part of this work revealed that tertiary amines with longer or branched alkyl chains allow tuning the phase stability and uranium extraction by modifying the volume of the polar species extracted in the organic phase. A complete fit of SANS data confirms in this second part that this phenomenon is related to the supramolecular self-assembly of the tertiary amines into smaller reverse micelle like aggregates for longer or branched alkyl chains. It moreover shows that these smaller aggregates quench third phase formation thanks to reduced attractive interactions between them. A thermodynamic analysis based on the “ienaic” approach further rationalizes this effect by showing that alkyl chains have a significant effect on the aggregate’s curvature free energy. The latter becomes a predominant inhibitor of uranium free energy of transfer due to the high entropic contribution originated by the extractants branching and packing.

Introduction

Solvent extraction is the method of choice applied in many industrial processes dedicated to metal recovery [1]. First studied empirically for historical applications, the knowledge of the mechanisms underlying solvent extraction processes are nowadays considered mandatory for their optimization [2].

Traditionally, the so-called “slope method” is applied to determine an apparent stoichiometry of the metallic species extracted in the organic solvent. Based on the mass action law, i.e., ignoring any volume effect in the polar cores-, the “slope method” used in chemical engineering leads to determination of the coordination number determination by the log–log plots of the extracted metal distribution coefficient vs the log of extractant concentration. The coordination number is defined as the number of extractants directly coordinated to each extracted cation, and is distinct from the aggregation number, that is the number of molecules par aggregate [3]. However, it has been noticed on several examples that the complex extraction mechanisms cannot be summarized with equilibria with well-defined stoichiometry [4], [5], [6]. For instance, deriving such complex stoichiometry from the variation of a distribution coefficient cannot account for the real structure of the organic phase.

Indeed, in addition to the strict extractant complexes necessary to chelate the cation to extract, it has been shown that the organic phase is composed of large amount of extractants in monomeric form, and weakly bounded extractant forming supramolecular aggregates that are comparable to reverse micelles [3]. The pseudo-phase approach introduced by Charles Tanford [7] applies here: monomers of extractant that are in the “bulk” pseudo-phase” are in dynamic equilibrium with aggregated extractants. Using the misleading “slope method” cannot account for all these effects.

As already mentioned in Part I, mechanisms underlying various features as acid and diluent effect, [8], [9], [10] third phase formation [11], [12], [13] or synergistic mixtures of extractants [14], [15] could be rationalized by taking into account complexation associated to extractant aggregation effects. It is not simple to disentangle chelation and aggregation mechanisms as they are usually coupled [16], [17]. Experimentally, one strategy is to modify parameters as extractant or diluent alkyl chains and measure all distribution coefficients to clearly isolate effect of aggregation from the extraction free energy. Studying lanthanide extraction by diglycolamides, Stamberga et al. [18] demonstrated with such a structure-performance relationship approach, that subtle steric changes markedly affect extraction and selectivity trends in lanthanides separation through electrostatic interactions occurring beyond the first coordination sphere of the metals extracted. This strategy was also followed in our study with tertiary amines extractants for uranium extraction, showing that alkyl chain length and branching significantly modify the extraction of polar species and the organic phase stability. This pure aggregation effect needs to be rationalized.

With a thermodynamic analysis Zemb et al., showed that considering the “motor” of complexation only is much too strong to explain the typical values of distribution coefficient obtained in solvent extraction [3], [19], [20]. In the colloidal approach called “ienaics”, [21] it was shown that a more general view going beyond molecular complexation considerations is necessary to account for the ion transfer from aqueous phase to organic phase. The Gibbs energy of ion transfer is decomposed in four terms, which take into account complexation as well as long range interactions in order to give insights on the origin of ion transfer.

The complexation with the first neighbors is a strong term of typically 50 kJ/mol per extracted species that allows the transfer to the oil phase [20]. This term is counter balanced by three weaker quenching terms:

- The “bulk” term or “droplet” term, corresponds to the free energy of a droplet of confined aqueous electrolyte solution in the core of the aggregates. This term models the entropic cost to pack the hydrophilic electrolytes inside the polar core of the aggregates. It can be calculated by considering the measured concentrations of water, acid and metals extracted in the organic phases.

- The “chain” term, is the sum of the curvature energy and of the micellisation energy of the aggregates:

- curvature energy represents the free energy of the extractants arrangement in the aggregates. It is associated with the variation of the shape and size of the reverse micelles which can be calculated from X-ray or neutrons scattering (aggregate and core radii), or from molecular mesoscopic simulations to provide the bending constant κ* and the spontaneous and effective packing parameters [17].

- micellisation free energy of the aggregates, can be calculated with the critical aggregation concentration (CAC) that can be derived from SAXS and SANS fitting or from surface tension measurements.

Based on these assumptions, Špadina et al. developed a theoretical model which allows quantitative estimation of free energies of ion transfer [17], [22], [23]. Successful application of “ienaic” approach have also been reported for neutral solvating extractant system, [22] for acidic extractant system, [17] as well as for synergistic extractants system [24].

We propose here to apply such a thermodynamic analysis on the tertiary amines used in the AMEX process for uranium production in the front-end nuclear fuel cycle. As in the first part of this work, extraction of two principal types of tertiary amines are considered to provide a complete description of the effect of alkyl chain length and branching on aggregation and extraction:

- a first group of 5 molecules with linear alkyl chains but different chains length: Trihexylamine (C6 THA), Tri-heptylamine (C7 THA), Tri-octylamine (C8 TOA), Tri-nonylamine (C9 TNA) and Tri-decylamine (C10 TDA);

- a second group of 3 molecules with the same carbon number on the alkyl chains but different branching situation: Tri-octylamine (C8 TOA), Tri-isooctylamine (C8 TIOA) and Tris(2-ethylhexyl)amine (C8 TEHA).

The two series of amines are presented in Fig. 1.

We showed in the first part of this study that tertiary amines with longer or more branched alkyl chains allow improving the phase stability and decrease uranium extraction when cumbersome branching are applied. These effects appeared to be related to smaller volume of the polar species extracted in the organic phase, as well as aggregates. To provide a deeper understanding on the alkyl chains configuration effect on extraction and aggregation and to decorrelate the thermodynamic motors responsible for such properties, we apply in this second part a complete thermodynamic analysis based on the “ienaic” approach and on the modeling of the scattering data.