Chemisorption of Na on the Ti Nanoparticle surface and its effects on the Na-H₂O reaction reactivity

This accident showed that elimination of SWR risk should be one of the most significant design criteria for the development of safe SFRs. Design solutions to ensure isolation of the Na from the water have been proposed; these include double-walled structures in the steam generator and at coolant boundaries and guard vessels or guard piping lines filled with inert gas. However, those methods cannot be ultimate solutions because the system still has a possibility of failure. An alternative approach to reduce the severity of the SWR is to reduce the reactivity of Na. Recently, this goal has been achieved by suspending a small amount of Ti nanoparticles (NPs) in liquid Na. Adding 2 at% of 10-nm Ti NPs in liquid Na reduced the heat of the reaction up to 80% and H₂ production rate by 10%. The number of adsorbed Na atoms onto Ti NPs with respect to the various temperature and pressure was calculated by using ab-initio and thermodynamics. Furthermore, the model which relates the Na chemisorption and SWR reactivity was proposed. The repression of SWR by Ti NPs was expected to be caused by strong chemical interaction between Ti NP surface and Na atoms; ab initio calculation using density functional theory and atomistic thermodynamic approach were conducted to test this hypothesis. Adsorption of 0.25ML Na at hcp hollow sites was the most favorable adsorption surface state and strong chemical adsorption was confirmed by the adsorption energy of -1.65 eV. A covalent-like metallic bond was confirmed between adsorbed Na and Ti atoms on the Ti(0001) surface; this bond affects the hydration energy of Na atoms and the activation barrier of the SWR. Using adsorption energy from the ab-initio calculation and atomistic thermodynamics, the thermodynamically most stable adsorption state at (T,P) was determined Na chemisorption effect on the SWR reactivity regression rate is modeled. The model predict χ variation with respect to time well especially dramatic reduction of χ; the χ at t < 1s from Saito,J-i et al experimental data within η_error=8% and Park,G et al experimental data within η_error=104%. Large discrepancy between the model and experiment results at Park,G et al, can be assumed to the error of C_NP,Interface evaluation