Vacuum insulation is applied in high voltage apparatus such as power circuit breakers and low loss capacitors. The highest possible electrical breakdown strength should be expected in ideal vacuum, since no charge carriers are present in the inter-electrode gap. Vacuum thus appears as an effective alternative to gas insulators such as sulfur hexafluoride (SF6), perfluorocarbons (CF4, C2F6, or the promising C4F8) which present the drawback of being global warming potential gases. But electrons emitted by the cathode directly cross the gap without any collision phenomena, and pre breakdown is frequently encountered, so limiting the use of vacuum for power circuit breakers. The role played by metallic microprotusions in the field emission mechanism is now well understood but an alternative emission mechanism proposed by Latham et al implies non-conductive or semi-conductive materials such as oxide layers or impurities, including adsorbed gases. Another field of application is in the controlled fusion domain. Fusion reactions in hot plasmas inside future Tokamak reactors (ITER, DEMO) are initiated by injecting high power beams of neutral D° atoms at high energy (1 MeV for ITER) into the plasma. Negative ions are accelerated by an intense electric field between electrodes at high voltage under vacuum and neutralised in a downstream gas target. The connection between the power supplies under SF6 and the electrodes under vacuum is made through an insulating passage called the bushing. Experiments at IRFM on the MV testbed (1MV, 100mA) have shown that the voltage holding is limited by the appearance of breakdowns if two electrodes are too close. The voltage holding with distance follows a square root law for distances larger than 1 cm. This dependence indicates according to the theory by Cranberg an exchange of micro clumps that cause breakdowns when they evaporate on the opposing electrode. Another performance limitation is the appearance of a sizeable electron current (100mA at 400kV) resulting from field emission that appears to follow the Fowler-Nordheim law. This unwanted dark current can be reduced, even eliminated, by the presence of gas in the vacuum vessel. This very beneficial feature is consistent with an increase of the work function of the metal. This would be caused by the adsorption of gas induced by an intense electric field. This physisorption process allows atoms to stick around emitting micro protrusions. Due to this, the emitting surface is reduced and the work function is increased. This research project aims to study the field-induced adsorption process by joining theoretical and modelling work with small-scale and large-scale experiments in several laboratories. The objective is to find physical conditions that favour the increase in surface work function, thus leading to an increased voltage holding in vacuum (by suppression of the dark current and absence of breakdowns) under high electric fields (50-100 kV/cm) between large electrode surfaces. This ANR project proposes specific research on the high-voltage holding under vacuum conditions. Four different laboratories work on 5 different themes: - Fundamental studies and modelling of field induced gas adsorption (LCAR Toulouse). - Simulation of field emission from a realistic surface and micro ionisation around emitting micro tips (LPGP, Paris). - Experimental study of the field emission and field-induced adsorption of gas with adjustable parameters (electrodes surface conditioning, electrodes material, electric field intensity, electrodes gap distance, gas nature and pressure); model validation (Supelec, Paris). - Study of the high voltage holding in vacuum using different materials and surface treatments to eliminate micro particles and micro tips. Large scale application of the previous more fundamental studies (IRFM, CEA Cadarache). - Construction and test of a prototype compact bushing, using all the knowledge gained in this project (IRFM, CEA-Cadarache).