In case of a severe accident on a nuclear reactor, radioactive iodine may be released into the environment, impacting significantly the radiological consequences. Determination of the amount released, and of the physical state of iodine (gaseous form or solid aerosol form), is thus a major issue. The release of iodine from the damaged reactor core and its transport in the different parts of the reactor up to the reactor containment, have been extensively studied, particularly in the Phébus-FP large scale experiments. Phébus-FP results notably showed that a significant fraction of iodine under gaseous form can reach the containment. The models used in severe accident codes did not (and still does not) fully account for this iodine speciation. A likely explanation is that iodine keeps a gaseous form up to the containment due to some processes that limit the formation of caesium iodide in the reactor coolant system (RCS) (caesium iodide was assumed to be the dominant form of iodine in the RCS). Caesium iodide formation would be limited due to chemical kinetic limitations and due to the presence of other elements (molybdenum or boron) responsible for "trapping" the caesium. An experimental research program has been developed with the aim to study the chemical behaviour of iodine during its transport in the RCS, with presence of steam, caesium and molybdenum or boron. The paper presents results obtained with molybdenum oxide. Experiments are compared to calculations performed with the IRSN severe accident code ASTEC where a chemical kinetic model has been implemented. The second part of this paper presents first results obtained with boric acid instead of molybdenum oxide.