In this paper, geometric bulk parameters, bulk moduli, energy gaps and relative stabilities of the TiO 2 anatase and rutile phases were determined from periodic DFT calculations. Then, for the rutile phase, structures, relaxations and surface energies of the (110), (100), (101) and (001) faces were computed. The calculated surface energies are consistent with the natural rutile powder composition, even if a dependence on the number of layers of the slab used to model the surface was identified. Internal constraints, consisting in freezing some internal 2 layers of the slab to atomic bulk positions, were thus added to mimic the bulk hardness in order to stabilize the computed surface energies for thinner systems. In parallel, the influence of pseudopotentials was studied and it appears that four valence electrons for titanium atoms are sufficient. The aim of this study was to optimise accurate rutile TiO 2 surfaces models that will be used in further calculations to investigate water and uranyl ion sorption mechanisms.

XPS and periodic DFT calculations have been used to investigate water sorption on the TiO 2 rutile (110) face. Two sets of XPS spectra were collected on the TiO 2 (110) single crystal 2 clean and previously exposed to water: the first set with photoelectrons collected in a direction parallel to the normal to the surface; and the second set with the sample tilted by 70°, respectively. This tilting procedure promotes the signals from surface species and reveals that the first hydration layer is strongly coordinated to the surface and also that, despite the fact that the spectra were recorded under ultra-high vacuum, water molecules subsist in upper hydration layers. In addition, periodic DFT calculations were performed to investigate the water adsorption process to determine if molecular and/or dissociative adsorption takes place.The first step of the theoretical part was the optimisation of a dry surface model and then the investigation of water adsorption. The calculated molecular water adsorption energies are consistent with previously published experimental data and it appears that even though it is slightly less stable, the dissociative water sorption can also take place. This assumption was considered, in a second step, on a larger surface model where molecular and dissociated water molecules were adsorbed together with different ratio. It was found that, due to hydrogen bonding stabilisation, molecular and dissociated water molecules can coexist on the surface if the ratio of dissociated water molecules is less than ≈ 33 %. These results are consistent with previous experimental works giving a 10-25 % range.

The sorption of Eu(III) onto kaolinite and montmorillonite was investigated up to 150°C. The clays were purified samples, saturated with Na in the case of montmorillonite. Batch experiments were conducted at 25, 40, 80 and 150°C in 0.5 M NaClO 4 solutions to measure the distribution coefficients (Kd) of Eu as a trace element (<10 À6 mol/L) between the solution and kaolinite. For the Na-montmorillonite, we used Kd results from a previous study [Tertre, E., Berger, G., Castet, S., Loubet, M., Giffaut, E., 2005. Experimental study of adsorption of Ni 2+ , Cs + and Ln 3+ onto Na-montmorillonite up to 150°C. Geochim. Cosmochim. Acta 69, 4937-4948] obtained under exactly the same conditions. The number and nature of the Eu species sorbed onto both clay minerals were investigated by time resolved laser fluorescence spectroscopy (TRLFS) in specific experiments in the same temperature range. We identified a unique inner-sphere complex linked to the aluminol sites in both clays, assumed to be @AlOEu 2+ at the edge of the particles, and a second exchangeable outersphere complex for montmorillonite, probably in an interlayer position. The Kd values were used to adjust the parameters of a surface complexation model (DLM: diffuse layer model) from 25 to 150°C. The number of Eu complexes and the stoichiometry of reactions were constrained by TRLFS. The acidity constants of the amphoteric aluminol sites were taken from another study [Tertre, E., Castet, S., Berger, G., Loubet, M., Giffaut, E. Acid/base surface chemistry of kaolinite and Na-montmorillonite at 25 and 60°C: experimental study and modelling. Geochim. Cosmochim. Acta, in press], which integrates the influence of the negative structural charge of clays on the acid/base properties of edge sites as a function of temperature and ionic strength. The results of the modelling show that the observed shift of the sorption edge towards low pH with increasing temperature results solely from the contribution of the @AlOEu 2+ edge complexes. Finally, we successfully tested the performance of our model by confronting the predictions with experimental Kd data. We used our own data obtained at lower ionic strength (previous study) or higher suspension density and higher starting concentration (TRLFS runs, this study), as well as published data from other experimental studies [Bradbury, M.H., Baeyens, B., 2002. Sorption of Eu on Na and Ca-montmorillonite: experimental investigations and modeling with cation exchange and surface complexation. Geochim. Cosmochim. Acta 66, 2325-2334; Kowal-Fouchard, A., 2002. Etude des mécanismes de rétention des ions U(IV) et Eu(III) sur les argiles: influence des silicates. Ph.D. Thesis, Université Paris Sud, France, 330p].

Periodic density functional theory (DFT) calculations using plane-wave basis sets were performed in order to study the bulk of nickel ferrite NiFe2O4. The local spin density approximation (LSDA) and the generalized gradient approximation (GGA) formalism were used, and it appeared that the LSDA failed to describe the magnetic structure of this compound. However, the GGA formalism gave reliable results in good agreement with experimental data for the lattice parameters, the electronic properties and the bulk modulus. In addition, the calculated density of states of the metallic species d block as well as their local magnetic moments were correlated to the crystal-field theory. Then, a charge deformation map was computed and, as expected from the electronegativity scale, the electron excess is localized around oxygen atoms along the bond axes. The formation energies of metallic vacancies are in good agreement with the inverse spinel structure experimentally observed.

Periodic density functional theory calculations have been performed in order to study the uranyl ion sorption on the TiO2 rutile (110) face. From experimental measurements, two uranyl surface complexes have been observed and the two corresponding sorption sites have been identified. However, from a crystallographic point of view, three different sorption sites can be considered on this face. The corresponding three surface bidentate complexes were modeled and optimized, and their relative energies were calculated. Only 5 kJ/mol separates the two most stable structures, which correspond to the experimental ones. The third surface complex is nearly 10 kJ/mol less stable, in agreement with the fact that it was not observed experimentally.