| It has become an international consensus that deep geological disposal is the most appropriate way for disposing high-level nuclear waste (HLW). The geological disposal is based on a multi-barrier concept which means the repository has barriers at three levels, i.e., waste package, engineered barrier and geological barrier (natural barrier). The clay minerals are important component in the engineered and geological barriers. The clay minerals notably present the interest to interact with most of the radionuclides, and the interaction/retention allows to slower the radionuclides transport. The retention property of clay minerals is therefore a key factor to study for the safety assessment of HLW repository. The quantitative prediction of the radionuclides retention requires the knowledge of the mechanisms occurring at the solid/liquid interface and their corresponding parameters. This thesis contributed to the development of sorption models for cations (Eu3+, Cs+ and Ni2+) on engineered and geological barriers. The source material used for representing the engineered barrier corresponds to Jinchuan Bentonite collected from Jinchuan, Gansu Province, China. For the geological barrier, the Callovo-Oxfordian formation (COx), a sedimentary clay-rock from Meuse/Haute Marne region (France) in which is supposed to build repository of HLW, was investigated. This thesis work was performed considering the following three open questions:(i) understanding of interaction occurring at solid/liquid interface at the atomic/molecular level,(ii) collecting and modelling data to complete the sorption database,(iii) applying known adsorption models developed in simple model systems to construct the model for the data collected in natural systems.It was a work performed in both Lanzhou University (Lanzhou, China) and laboratory Subatech (Nantes, France) in the frame of two financed programs, i.e. "Adsorption, desorption and transfer of Actinides in the presence of phosphate and humic acid" from the National Natural Science Foundation of China and the "groupement de laboratoire transfert" from National Radioactive Waste Management Agency (Andra). The manuscript was divided into 5 chapters. The chapter 1 described the context of the work, the open questions and the aims of this study. The chapter 2 presented the study of Eu(â…¢) adsorption on bentonite in the presence of phosphate. The following chapters 3-4 presented the Cs+ and Ni2+ adsorption on COx and corresponding modelling work. The chapter 5 was a molecular dynamic study of Cs on interstratified illite smectite (I/S) to understand one question we find in chapters 3, i.e., whether I/S can be considered as a mechanical mixture of illite and semectite. The General conclusions and perspectives were summarized in chapter 6.The results obtained in this thesis were briefly given below.1. Adsorption of Eu(III) on bentonite in the presence of phosphateThe adsorption of phosphate on bentonite slightly increased in the pH range of 2.5-6.5, and decreased in the pH range of 6.5-9.4. This adsorption can be quantitatively interpreted by a model considering the formation of three monodentate surface complexes,=SH2PO4,=SHPO4-, and =SPO42, or the formation of two monodentate surface complexes and one bidentate surface complex,=SH2PO4,=S2HPO4 and=SPO42-.As the first model gave a slightly better description of the experimental data, it was selected as the sub-model to describe the ternary system. In the ternary system, a synergistic adsorption was observed in the presence of both phosphate and Eu(â…¢). In addition to the two sub-models describing Eu(III) and phosphate adsorption, the formation of ternary surface complexes had to be considered in order to explain the synergistic effect experimentally observed. The experimental data could be quantitatively explained when Eu(â…¢) (=SOEuH2PO4+ and=SOEuHPO4) or phosphate (=SPO4Eu+) are the bridged atoms. Complementary experiments carried out by X-ray photoelectron spectroscopy suggested that the second case is the most probable. The proposed model can be used in order to predict Eu(â…¢) adsorption on buffer/backfilling material in the presence of phosphate, which originates from the radiolysis of phosphorous-containing extractants.2. Cs(â… ) adsorption on Callovo-Oxfordian clay-rockCs(â… ) sorption is dominated by the clay fraction of the COx and the mechanism is reversible in the experimental conditions of this study. The retention can be quantitatively described based on a bottom-up approach with two reference clay minerals (illite and montmorillonite) using published parameters. For the interstratified illite-smectite, only the planar sites were must be considered, thereby demonstrating that, in general, the sorption properties of the mixed layered minerals cannot be considered as an average of the sorption properties of their constituents. The proposed model can explain the sorption of Cs by COx for a wide range of samples of variable mineralogy. It can also predict the effect of slight changes in the composition of the expected water in the formation and explain Cs retention on compacted/intact COx samples as well as the behavior of natural Cs. No change in the cation exchange capacity occurs between the dispersed and compacted states (i) in agreement with results reported for the Opalinus Clay formation that is richer in illite but poorer in I/S mixed layer than the COx formation but (ii) at variance with the results reported for highly compacted montmorillonite. Furthermore, In all the cases studied (reversibility tests performed after a sorption phase in the compacted state fixed between 5 months and 5 years), the sorption process was shown to be reversible in disagreement with results obtained on similar but different illite materials, stressing the need to study not only generic clay samples but also samples that are relevant to the sites under consideration. The transposition of the model to the scale of the COx formation indicates a significant retention of Cs by the clay formation with Kd values ranging from 100 to 2000 L/kg for trace concentrations of Cs. This retention property makes the geological barrier efficient for preventing Cs migration if it is released from the repositoryfrom the storage facility.3. Ni(II) adsorption on allovo-Oxfordian clay-rockThe "bottom-up" approach considering the clay fraction as a mixture of illite and montmorrilonite was appropriate for describing the cation exchange process on the clay-enriched <2μm fraction with published parameters. However, a less good agreement was obtained between prediction and experimental results when the surface complexation mechanism considered within the 2 SPNE/SC model is used for describing Ni sorption on COX. An operational approach based on the generalized composite model (GCM) was therefore developed for describing Ni sorption on COX. It corresponds basically to the model proposed for the Opalinus Clay omitting the strong site and in which the surface complexation constants for the weak site reactivity strength haves been adjusted to the experimental data without linking the reactivity strength with a pure model clay phase slightly increased. This difference in affinity cannot be explained by a competition effect but might be related to a difference in reactivity between the COX edge sites. Integrating this operational model into an additive approach taking into account clay mineralogy allows prediction of the behavior of Ni(II) on the representative a COX sample having a relative high clay fraction (-50%). The crucial role of this clay fraction in the retention was confirmed at the molecular level by XPS and EXAFS. However, the model underestimates the retention when the clay content decreases, i.e. when the content in carbonate phases increases. It notably and largely underestimates the sorption for the sample with the highest carbonate content (85%). Further experimental data will be necessary to understand the origin of this extra sorption. A Kd range from 60 to 300 L/kg for trace concentrations of Ni(II) (～5.10-87-10-5 M) is appropriate for clay-rock having clay contents above 1.5%.4. A molecular simulation study for Cs+-I/S systemBy performing MD simulations and analyzing the resulting trajectories, it was shown that Cs+ forms only inner-sphere complex on the I/S basal surface and Na+ can form both inner-sphere and out-sphere complex. There is only one sorption site for Cs+ (hexagonal cavity center), whereas there are two inner-sphere sorption sites for Na+ (hexagonal cavity center and tetrahedral site, Figure below). Cs+ and Na+ can present together on the I/S basal surface (at different adsorption site). Since Cs+may not be sorbed on the tetrahedral site, Cs+â†'Na+ cation exchange reaction on the I/S surface corresponds to the reaction on the hexagonal cavity center only. This result may indicate the main mechanism of Cs+adsorption on Na+-I/S basal surface. However, the real system is more complicated, so other possibilities (i.e. Cs+adsorption on hexagonal site is accompanied with Na+ desorption from tetrahedral site) may co-exist in this adsorption procedure. |
PDF Full Text Request