Reactions of Sodium-Potassium Alloys with Inert Gas Impurities − Potential Hazards after Oxidation

Abstract

A bibliographical study of the reactions of NaK alloys with their oxides did not explain the related accidents that have occurred in the past, particularly those involving the handling of bubblers installed to remove traces of oxygen and water vapor from blanket gases of LMFR (liquid metal fast reactors). Moreover, it revealed a controversy with respect to the explosive hazard of the NaK alloys + KO2 reaction. A thermodynamic study of the reaction of a NaK alloy with oxygen as well as experimental work on mixtures containing NaK and different oxides (KO2, Na2O, Na2O2) led to the conclusion that violent reactions occurring at ambient temperature, were always linked to the presence of hydrated superoxides or hydroxides. On contact with NaK alloys, these hydrated compounds lead to the formation of hydrogen with a release of heat causing the decomposition of potassium superoxide KO2 and the simultaneous release of oxygen. The oxidation of liquid NaK alloy, by bubbling a mixture of argon and air containing traces of water vapor (H2O < 5 vpm), leads to the formation of mainly KO2, besides the hydrated compounds Na2O2·2H2O, NaOH·H2O and KOH·H2O. The gas follows preferential paths and the product obtained is highly heterogeneous in both composition and hardness (porous parts and dense parts). Hydrated superoxide and hydroxides treated under vacuum result in nearly quantitative dehydration after heating at 100 °C. For storage elements or safety valves, containing NaK that has been submitted to slow surface oxidation and hydration, the oxidized layer contains KO2, Na2O2·nH2O and hydrated hydroxides of sodium and potassium. In order to avoid any hazard due to the contact of the oxides with the residual NaK alloy during transport or under impact, it is necessary to freeze the alloy and to reheat it very slowly. Operations such as draining should also be performed slowly, after having calibrated the tanks (temperature and pressure). Therefore the use of NaK liquid alloy to purify gases is questionable; as preferential paths are established, it remains impossible to predict the period of efficiency of a bubbler. Moreover, the NaK alloy does not chemically transform all traces of water vapor; because some remains present in the form of hydrated products, leading to the formation of wastes with a potential hazard during storage or destruction. In our opinion, this method of purification should be abandoned.