1. Introduction In spite of the potential to mitigate the global warming and replace the current fossil fuel power generation, nuclear power generation has a problem on the geological disposal of radioactive wastes from the viewpoint of long-term safety. Thus, a new disposal process is proposed in Japan that the long-lived fission products (LLFPs) such as 135Cs, 79Se, 93Zr, and 107Pd whose half-lifes are around million years are separated and converted into short-lived or stable nuclides by nuclear transmutation [1]. According to the conventional scenario of geological disposal, huge amount of vitrified wastes have been produced. Since LLFPs are immobilized in a three-dimensional (3-D) network structure of glass in the vitrified wastes, it is necessary to develop a new method to recover LLFPs from the vitrified wastes. As a first-step of the recovery method, we propose the electrochemical reduction in molten CaCl2 to break the 3-D network structure of borosilicate glass in the vitrified wastes. In the present study, as an initial consideration, we investigated the electrochemical reduction behavior of borosilicate glass without LLFPs in molten CaCl2. 2. Experimental The experiments were conducted in molten CaCl2 in a dry Ar atmosphere at 1123 K. As a working electrode, a glass-seal electrode which was prepared by sealing a tungsten rod (diameter: 2.0 mm) into an alkali borosilicate glass tube (Pyrex®, SiO2 81.1 wt%, B2O3 12.6 wt%, Al2O3 2.3 wt%, Na2O/K2O 4.0 wt%, o.d. 8 mm) was used. The counter and reference electrodes were a graphite rod and an Ag+/Ag electrode, respectively. After potentiostatic electrolysis at 0.9 V vs. Ca2+/Ca for 30 minutes, the glass-seal electrodes were rinsed by distilled water, and then characterized by SEM/EDX and XRD. For XPS measurement, the samples were consecutively rinsed by distilled water, 8 % HCl aqueous solution, and 10 % NaOH aqueous solution. 3. Results and Discussion 3-1. Thermodynamic Analysis & Cyclic Voltammetr y According to the Gibbs energies of formation for the oxides composing borosilicate glass (MO x , M = Si, B, Al, Na, K), K2O, Na2O, and B2O3 are less stable than SiO2 at the experimental temperature of 1123 K. On the other hand, the reduction of Al2O3 occurs at more negative potentials than the reduction of SiO2. From the Gibbs energies of formation for the related oxides and chlorides (MO x and MCl x , M = Ca, Si, B, Al, Na, K), the following reactions of the oxides with molten CaCl2 are thermodynamically favorable. Na2O (s) + CaCl2 (l) = CaO (s) + 2 NaCl (l) ΔGº(1123K) = −246.5 kJ mol-1 K2O (s) + CaCl2 (l) = CaO (s) + 2 KCl (l) ΔGº(1123K) = −347.7 kJ mol-1 Namely, the dissolutions of Na2O and K2O from borosilicate glass are highly expected only by immersion in molten CaCl2. In the cyclic voltammogram for a borosilicate glass-seal electrode, cathodic current was observed at more positive potential than that for silica glass [2]. Based on the thermodynamic calculations, the reduction of B2O3 was suggested to occur in this potential region. 3-2. Characterization of Products Prepared by Potentiostatic Electrolysis An XRD pattern of the reduced borosilicate glass confirmed the reduction of SiO2 and the formation of crystalline silicon. From XPS spectra for B 1s before and after the reduction, the weakened B–O peak and strengthened B–B or B–Si peak were clearly observed. The existence of the reduced state of boron confirms the reduction of B2O3. The reduction of SiO2 was also confirmed by XPS Si 2p spectra. As shown in the SEM image in Fig. 1, granules with a diameter of 1~10 μm were produced in the reduction product. From the EDX mapping results, the oxide portions in the sample always contain all of Al, Ca, and Cl. This indicates that aluminum oxide is not reduced at 0.9 V, and that calcium aluminate is formed by reacting with CaO in the molten salt. The existence of chlorine is explained by the formation of complex compounds consisting of calcium aluminate and CaCl2. The dissolution of Na2O into molten CaCl2 was also confirmed from EDX analysis. Acknowledgement This work was partly funded by ImPACT Program of Council for Science, Technology and Innovation (Cabinet Office, Government of Japan).
Read full abstract