Validating a model to predict the chemistry of the fuel used in radioisotope power systems

U.S. radioisotope power systems convert the decay heat from
²³⁸
PuO
₂
fuel into electricity using various forms of thermoelectric conversion technology. When placed in an environment that has low oxygen potential and very high temperatures, the fuel can undergo reduction to a sub-stoichiometric form (PuO
₂
-x). In order to better understand this system, a model was developed that combines the known relationship between stoichiometry (i.e., x), temperature, and oxygen potential with the chemical thermodynamics of graphite, carbon monoxide, and carbon dioxide. Because of the challenges surrounding PuO
₂
experimentation, the model has been benchmarked against a common PuO
₂
surrogate: CeO
₂
. Samples of CeO
₂
were sealed into a reaction tube and allowed to react with graphite using a reaction temperature of 1273 K. When these reaction conditions were used, the model (x = 0.268) and experiment (x = 0.272 + 0.008) found very good agreement. Successful benchmarking of this model suggests that it will be useful in predicting the complex chemical interaction that can occur between the materials inside a radioisotope power system. Kinetic data obtained from these experiments suggests that reaction rates are very slow and that it could take hundreds of hours to reach equilibrium. A change in the rate limiting reaction mechanism is also observed when x > 0.25, but does not appear to impact the thermodynamic equilibrium. In addition to helping inform a
²³⁸
Pu based system, this model may also be useful for predicting the chemistry found inside an
²⁴¹
Am based system, which is currently under development for the European Space Agency.