Structure, equilibrium and ligand exchange dynamics in the binary and ternary dioxouranium(vi)-ethylenediamine-N,N′-diacetic acid-fluoride system: a potentiometric, NMR and X-ray crystallographic study

Abstract

The structure, thermodynamics and kinetics of the binary and ternary uranium(VI)-ethylenediamine-N,N′-diacetate (in the following denoted EDDA) fluoride systems have been studied using potentiometry, ¹H, ¹⁹F NMR spectroscopy and X-ray diffraction. The UO₂²⁺–EDDA system could be studied up to −log[H₃O⁺] = 3.4 where the formation of two binary complexes UO₂(EDDA)(aq) and UO₂(H₃EDDA)³⁺ were identified, with equilibrium constants logβ(UO₂EDDA) = 11.63 ± 0.02 and logβ(UO₂H₃EDDA³⁺) = 1.77 ± 0.04, respectively. In the ternary system the complexes UO₂(EDDA)F⁻, UO₂(EDDA)(OH)⁻ and (UO₂)₂(µ-OH)₂(HEDDA)₂F₂(aq) were identified; the latter through ¹⁹F NMR. ¹H NMR spectra indicate that the EDDA ligand is chelate bonded in UO₂(EDDA)(aq), UO₂(EDDA)F⁻ and UO₂(EDDA)(OH)⁻ while only one carboxylate group is coordinated in UO₂(H₃EDDA)³⁺. The rate and mechanism of the fluoride exchange between UO₂(EDDA)F⁻ and free fluoride was studied by ¹⁹F NMR spectroscopy. Three reactions contribute to the exchange; (i) site exchange between UO₂(EDDA)F⁻ and free fluoride without any net chemical exchange, (ii) replacement of the coordinated fluoride with OH⁻ and (iii) the self dissociation of the coordinated fluoride forming UO₂(EDDA)(aq); these reactions seem to follow associative mechanisms. ¹H NMR spectra show that the exchange between the free and chelate bonded EDDA is slow and consists of several steps, protonation/deprotonation and chelate ring opening/ring closure, the mechanism cannot be elucidated from the available data. The structure (UO₂)₂(EDDA)₂(µ-H₂EDDA) was determined by single crystal X-ray diffraction and contains two UO₂(EDDA) units with tetracoordinated EDDA linked by H₂EDDA in the “zwitterion” form, coordinated through a single carboxylate oxygen from each end to the two uranium atoms. The geometry of the complexes indicates that there is no geometric constraint for an associative ligand substitution mechanism.