Remobilization of fission products (FPs) impacting the delayed source Term

The ASTEC code is the integral code used to perform the simulation of all phenomena occurring during a reactor meltdown accident, from the initiating event up to the release into the environment. The ASTEC simulation of the FPs behavior have shown some discrepancies in the quantitative prediction of observations in the Benchmark Study of the Accident at the Fukushima Daiichi Nuclear Power Station (BSAF), as well as some detailed interpretations of Phebus FP and CHIP experiments. These works have highlighted to consider more in detail the reactions of the FPs of interest with aerosols and surfaces (which promote their trapping in the facilities), as well as their remobilization from the surfaces into the gaseous phase, and thus their contribution to a potential release into the environment in the medium and long terms. Some experimental works have been initiated in the ESTER project and theoretical chemistry approaches have been developed to explore atomic-scale mechanisms between the surfaces and gases.

Molybdenum oxide is of particular interest because Mo, having a high fission yield, can interact with some other fission products (e.g. Cs), favoring in some conditions the gaseous iodine fraction. Thus, the main objective of the thesis is to improve the ASTEC modelling by providing thermodynamic and kinetic data about the surface reactivity of molybdenum oxide.

To achieve this goal, an extensive ab-initio study of molybdenum trioxide surface reactivity deposited in the primary circuit using Density Functional Theory was performed. We first focused our study on two phenomena: activated dissociation of CsI and non-congruent condensation of CsHMoO₄ , both important species for the iodine transport simulation.

First, the Gibbs energies of adsorption and dissociation of CsI on O- and Mo-terminated surfaces are determined as a function of temperature. As result, Mo-surface doesn't activate CsI dissociation to form I₂ , whereas O-terminated surface does because the Gibbs energies of adsorption and dissociation occur in the same temperature range of 350–550 K. These results are consistent with Bader charge analysis, in which the I on the O-terminated surfaces show higher charge loss compared to the Mo-terminated surface case, confirming that CsI dissociation is more likely to occur on the O-terminated surface and, therefore, favoring the interaction with a neighboring iodine.

Concerning CsHMoO₄ , it is found that the non-congruent condensation on both surfaces is unlikely, since the adsorption and dissociation are not thermodynamically possible at the same temperature intervals.