The structures of the UO2(aq)2+ ion and of the uranium(VI) hydroxide complex(es) formed in strongly alkaline solution have been investigated theoretically using molecular-orbital based quantum chemical methods, and experimentally using EXAFS methodology. Relativity was included explicitly through the Douglas−Kroll transformation. The uranium atom was described at the ECP level, using the AIMP methodology. The structures of [UO2(H2O)5]2+, and the hydroxide complexes, viz., [UO2(OH)4·(H2O)]2-, [UO2(OH)4]2-·(H2O), [UO2(O)(OH)2]2-·2(H2O), and [UO2(OH)5]3-, were optimized at the SCF level, using gradient techniques, while the relative stabilities were calculated at the MP2 level of approximation. The third structure contains three coordinated ligands, one of which is an oxide ion, in the plane perpendicular to the linear UO2-unit. Complexes of this type have not been experimentally identified for U(VI); however, they are formed for the iso-electronic Np(VII). The experimental EXAFS data indicates that the complex(es) formed is(are) mononuclear. The number of coordinated ligands in the equatorial plane is 4.5 ± 0.4, while the bond distances are the same within the experimental errors, as in a previous study of [Co(NH3)6 3+]2[UO2(OH)4 2-]3·2H2O, by Clark et al. An EXAFS model where the coordination number is fixed to four, is only marginally less precise than the model without constraints on the coordination number. This fact together with the close agreement between experimental and theoretically observed variations in bond distances between the different structure models provides a strong indication for the formation of [UO2(OH)4]2- in solution. This is an unusual coordination number for uranium(VI) complexes, previously found in sterically crowded systems such as UO2Cl4 2-.

The structures of the complexes UO(2)F(n)(H(2)O)(5-n)(2-n), n = 3-5, have been studied by EXAFS. All have pentagonal bipyramid geometry with U-F of and U-H(2)O distances equal to 2.26 and 2.48 A, respectively. On the other hand the complex UO(2)(OH)(4)(2-) has a square bipyramid geometry both in the solid state and in solution. The structures of hydroxide and fluoride complexes have also been investigated with wave function based and DFT methods in order to explore the possible reasons for the observed structural differences. These studies include models that describe the solvent by using a discrete second coordination sphere, a model with a spherical, or shape-adapted cavity in a conductor-like polarizable continuum medium (CPCM), or a combination of the two. Solvent effects were shown to give the main contribution to the observed structure variations between the uranium(VI) tetrahydroxide and the tetrafluoride complexes. Without a solvent model both UO(2)(OH)(4)(H(2)O)(2-) and UO(2)F(4)(H(2)O)(2-) have the same square bipyramid geometry, with the water molecule located at a distance of more than 4 A from uranium and with a charge distribution that is very near identical in the two complexes. Of the models tested, only the CPCM ones are able to describe the experimentally observed square and pentagonal bipyramid geometry in the tetrahydroxide and tetrafluoride complexes. The geometry and the relative energy of different isomers of UO(2)F(3)(H(2)O)(2-) are very similar, indicating that they are present in comparable amounts in solution. All calculated bond distances are in good agreement with the experimental observations, provided that a proper model of the solvent is used.

To assess the nature of chemical bonds in uranyl(VI) complexes with Lewis base ligands, such as F(-), Cl(-), OH(-), CO(3)(2-), and O(2)(2-), we have used quantum chemical observables, such as the bond distances, the internal symmetric/asymmetric uranyl stretch frequencies, and the electron density with its topology analyzed using the quantum theory of atoms-in-molecules. This analysis confirms that complex formation induces a weakening of the uranium-axial oxygen bond, reflected by the longer U-O(yl) bond distance and reduced uranyl-stretching frequencies. The strength of the ligand-induced effect increases in the order H(2)O < Cl(-) < F(-) < OH(-) < CO(3)(2-) < O(2)(2-). In-depth analysis reveals that the trend across the series does not always reflect an increasing covalent character of the uranyl-ligand bond. By using a point-charge model for the uranyl tetra-fluoride and tetra-chloride complexes, we show that a significant part of the uranyl bond destabilization arises from purely electrostatic interactions, the remaining part corresponding either to charge-transfer from the negatively charged ligands to the uranyl unit or a covalent interaction. The charge-transfer and the covalent interaction are qualitatively different due to the absence of a charge build up in the uranyl-halide bond region in the latter case. In all the charged complexes, the uranyl-ligand bond is best described as an ionic interaction. However, there are covalent contributions in the very stable peroxide complex and, to some extent, also in the carbonate complex. This study demonstrates that it is possible to describe the nature of chemical bond by observables rather than by ad hoc quantities such as atomic populations or molecular orbitals.

The mechanisms for the exchange of water between [UO(2)(H(2)O)(5)](2+), [UO(2)(oxalate)(2)(H(2)O)](2)(-)(,) and water solvent along dissociative (D), associative (A) and interchange (I) pathways have been investigated with quantum chemical methods. The choice of exchange mechanism is based on the computed activation energy and the geometry of the identified transition states and intermediates. These quantities were calculated both in the gas phase and with a polarizable continuum model for the solvent. There is a significant and predictable difference between the activation energy of the gas phase and solvent models: the energy barrier for the D-mechanism increases in the solvent as compared to the gas phase, while it decreases for the A- and I-mechanisms. The calculated activation energy, Delta U(++), for the water exchange in [UO(2)(H(2)O)(5)](2+) is 74, 19, and 21 kJ/mol, respectively, for the D-, A-, and I-mechanisms in the solvent, as compared to the experimental value Delta H(++) = 26 +/- 1 kJ/mol. This indicates that the D-mechanism for this system can be ruled out. The energy barrier between the intermediates and the transition states is small, indicating a lifetime for the intermediate approximately 10(-10) s, making it very difficult to distinguish between the A- and I-mechanisms experimentally. There is no direct experimental information on the rate and mechanism of water exchange in [UO(2)(oxalate)(2)(H(2)O)](2-) containing two bidentate oxalate ions. The activation energy and the geometry of transition states and intermediates along the D-, A-, and I-pathways were calculated both in the gas phase and in a water solvent model, using a single-point MP2 calculation with the gas phase geometry. The activation energy, Delta U(++), in the solvent for the D-, A-, and I-mechanisms is 56, 12, and 53 kJ/mol, respectively. This indicates that the water exchange follows an associative reaction mechanism. The geometry of the A- and I-transition states for both [UO(2)(H(2)O)(5)](2+) and [UO(2)(oxalate)(2)(H(2)O)](2-) indicates that the entering/leaving water molecules are located outside the plane formed by the spectator ligands.