Disruptions are abrupt interruptions of plasma discharges in tokamaks. They are due to instabilities leading to the loss of thermal energy and magnetic energy of the plasma over periods of the order of a few tens of milliseconds. Disruptions can generate so-called relativistic runaway electron beams reaching energies up to several MeV and potentially carrying a large part of the initial current. It is crucial to control or stop them to ensure a reliable operation of future tokamaks such as ITER. The proposed thesis project focuses on the mitigation of runaway electrons by massive injection of deuterium or hydrogen into the beam. This scenario leads to a drastic decrease in the energy deposited on the wall by the runaway electrons, through two phenomena: a magnetohydrodynamic instability and the absence of regeneration of the runaway electrons in the final loss of the plasma current. These two conditions are obtained when the plasma created by the interaction between the runaway electron beam and the neutral gas remains cold enough to recombine. The recombination mechanism relies on energy transport processes by the neutrals and a decrease in the interaction between the runaway electrons and the background plasma. Limits to this scheme were found on current tokamaks; they must be understood in order to extrapolate to future machines. The first part of the thesis will focus on the characterization of the cold plasma: density profiles, deuterium/hydrogen or heavy impurity concentration, current profile. We will be particularly interested in the quantities related to transport phenomena in the plasma: heat conduction, particle diffusion or radiation transport. This experimental characterization will quickly call upon numerical modelling to confirm the role of the various transport mechanisms in keeping the conditions required for the dissipation of the beam without damage. An extrapolation towards ITER will then be considered via simulations.