Characterization of secondary phases of spent nuclear fuel under final geological disposal conditionsTheoretical and experimental studies

  1. Colmenero Ruiz, Francisco
Dirigida por:
  1. Vicente Timón Salinero Director/a
  2. Joaquín Cobos Sabaté Director/a

Universidad de defensa: Universidad Autónoma de Madrid

Fecha de defensa: 21 de septiembre de 2017

Tribunal:
  1. Rafael Escribano Torres Presidente/a
  2. Manuel Yáñez Montero Secretario/a
  3. Valentín García Baonza Vocal
  4. Valentín Gómez Escobar Vocal

Tipo: Tesis

Teseo: 507253 DIALNET

Resumen

The resurgence of the use of nuclear energy is mainly driven by the need of more electricity for increasing populations and consumption, as well as the need of energy sources without CO2 emissions and other greenhouse gases causing global warming. However, the management of nuclear waste is a matter of concern because it can be a source of high ecological damage, in the same way as uranium ore mining and nuclear accidents. Nuclear fuel, commonly composed of UO2 enriched from 0.7 to 3–5 % of 235U, is obtained from natural minerals found in the rocks of the Earth´s crust. After its irradiation (one or more irradiation cycles) the fuel is considered as spent nuclear fuel (SNF) and must be managed as waste. The SNF is composed of a UO2 matrix (> 95 %) and other radioactive elements. The latter are very hazardous, making the waste management difficult. Their hazard progressively decreases by natural processes (radioactive decay) leading, after several millions of years, to a total radioactivity that equals the radioactivity of natural uranium. Therefore, it has been proposed that the most appropriate and natural way of managing this waste is to return it to the Earth’s crust. For this aim, the generally accepted solution is the burial of the SNF in a deep geological disposal for a period, at least, as long as the radioactive decay time. The design of a deep geological disposal must avoid the reaching of these radionuclides to the biosphere for such a long time by using a multi-barrier system. In this system, the radioactive waste is confined inside cladding tubes and canisters, which, in turn, are protected by introducing them into tunnels filled with buffer material (e.g. bentonite) and, finally, the whole system is surrounded by geological natural barriers. However, it is well-known that, after such a long time the barriers that protect the waste will be breached and SNF will be in contact with water. Thus, water could be the vehicle that mobilizes radionuclides. Although the groundwater conditions in a repository are generally reducing, in a layer near the fuel surface, with a thickness < 50 µm, an oxidative environment has been postulated. This ambient is produced by the radiolysis of water due to the ionizing radiation associated with the fuel. The radiolysis of groundwater results in the production of oxidants as H2O2 among others. Therefore, uranium in the matrix of the spent nuclear fuel, composed by uranium dioxide, UIVO2, could oxidize to U(VI) and dissolve into the water forming uranyl groups. These uranyl groups can precipitate forming different secondary phases, i.e. alteration products, on the spent fuel surface depending on the local conditions and concentrations of reactive species present. Since uranyl compounds, containing UO22+, will appear as secondary phases at the surface of spent nuclear fuel in the final geological disposal conditions, its characterization and the determination of its properties are extraordinarily important. However, due to the radiotoxicity of these materials, the experimental study of these substances requires an extremely careful handling of the samples. Theoretical methods, based in the laws of quantum physics and chemistry studying these systems at atomistic level, are free of these difficulties and have been used in this work both as an interpretative tool of the experimental results and as a predictive tool for the properties of these substances, inaccessible through their experimental measurement. The uranyl minerals rutherfordine, studtite, soddyite and uranophane, which have been widely recognized to be fundamental components of the paragenetic sequence of secondary phases that arises from the weathering of uraninite ore deposits and corrosion of SNF, have been studied by means of X-Ray powder diffraction and Raman spectroscopy combined with first principle calculations based on density functional theory (DFT) methods. In the initial steps of this study, it was found that within scientific bibliography there were only a very few studies on the theoretical vibrational spectra of these materials. In fact, none of these initial studies included a complete determination of these spectra since only the phonon frequencies at gamma point were calculated and band intensities were not reported. The reasons for the very limited amount of theoretical treatments were very clear once the bibliography was revised: 1) the high level of theory required to describe the uranium atom containing systems; 2) the large number of atoms in the unit cells of these materials; 3) the large number of valence electrons which must be described explicitly in the calculations; and 4) the absence of good norm-conserving pseudopotentials needed for the computation of the theoretical vibrational spectra. Having these difficulties in mind, it was thought that the best way to deal with these problems was the generation of a completely new pseudopotential for uranium atom. The pseudopotential should satisfy, at least, the conditions of being relativistic and norm-conserving if we desire to study the structure and spectra of these compounds with a good level of accuracy. The pseudopotential was generated and its performance was evaluated for a large series of uranium-containing minerals. The structure of these materials was determined and the resulting lattice parameters, bond lengths, bond angles and X-Ray powder diffraction pattern were found in very good agreement with experimental data. Since the structures of many of these materials (for example, soddyite and uranophane) were never studied by means of rigorous theoretical solid-state calculations, our work confirmed by the first time the structures determined experimentally by means of the X-Ray diffraction technique. The Raman spectrum of the selected uranyl minerals (rutherfordine, studtite, soddyite and uranophane) was experimentally determined at CIEMAT and calculated theoretically using density functional perturbation theory (DFPT) at CSIC. A detailed comparison of the theoretical and experimental spectra is reported in this work. The non-scaled wave numbers of the bands of the theoretical spectrum showed very good agreement with the corresponding experimental values. This concordance in the comparison allowed for the rigorous assignment of the bands in the Raman spectra, since the theoretical methods provide microscopic scale views of the atomic motions in the corresponding normal modes. These spectra were the first published complete theoretical Raman spectra (including band wavenumbers and intensities) of uranyl-containing crystalline materials. Since the calculations provided very good quality structures and spectra, the possibility of obtaining additional properties of these materials was considered. Due to the importance of the structures of these materials, their mechanical stability was studied. The mechanical stability and properties, as well as the equations of state of the uranyl carbonate mineral rutherfordine, and the uranyl silicate minerals soddyite and uranophane were studied theoretically for the first time. Thermodynamic properties, indispensable for the dynamic modelling of the chemical behavior of these compounds, were also determined. For the case of rutherfordine the results of the theoretical calculations of these properties showed an excellent agreement with experimental data. This fact suggest that theoretical methods constitute an excellent tool, alternative to experimental methods, for the determination of the thermodynamic properties of uranium containing materials. A side result from the rutherfordine mineral study is that our results extended the temperature range in which the thermodynamic properties of this material were known to the range from 0 to 700 K. The good quality of the thermodynamic properties calculated theoretically has been confirmed by the reliable results obtained in a later work of gamma uranium trioxide, γ-UO3. By using the calculated structures and Raman spectra of a set of uranyl containing materials, an UO bond-length to uranyl symmetric stretching Raman shift relation was obtained by a numerical fit of the calculated data. The relation, when used to determine UO bond-lengths from experimentally determined Raman shifts, provided bond-lengths of similar accuracy than other empirical relations widely used in the literature obtained by a fit to a large set of data points deduced from X-Ray diffraction and Raman spectroscopic measurements. In contrast, the relation from this work was obtained from a small set of data points deduced from first principle calculations, that is, without using any empirical information. Because of secondary phases of SNF can reduce the release of fission products and heavier actinides contents in the SNF to the biosphere, the possibility of incorporation of fission products (the main hazard of SNF if the Advanced Nuclear Fuel Cycle is adopted) into the crystal structure of uranophane-α was also studied. It was found that the incorporation of strontium into the structure of this mineral was possible. The Sr-exchanged solid material has an X-Ray powder diffraction pattern and a Raman spectrum very close to that of normal uranophane in agreement with the results of previous experimental studies. This study demonstrates that this uranyl silicate mineral may act as an additional barrier for the release of fission products to the biosphere. In summary, the theoretical results, including the determination of structural, mechanical, thermodynamic and vibrational properties, suggest that theoretical methodology is a very powerful tool to perform detailed studies of the secondary phases of SNF free of the difficulties associated to its experimental handling due to their radiotoxicity. Finally, it can be concluded that the interplay of experimental and theoretical methods enhances extraordinarily the knowledge and understanding about the processes and compounds involved in the disposal of spent nuclear fuel.