Negative ion sources for neutral beam injection in fusion experiments are based on the surface production of Hor Don caesiated low work function surfaces. In the last years it was demonstrated at the large RF driven ion source of the ELISE test facility that the requirements for the ITER NBI systems can be fulfilled in hydrogen. This is a big step towards the first operational period of ITER, planned for up to 2035. However, for the following operational period neutral beam systems working in deuterium are needed. Operation of negative hydrogen ion sources in deuterium is significantly more demanding than in hydrogen: the amount of co-extracted electrons is much higher and their increase during pulses is much more pronounced, limiting the achievable performance. This paper presents results of investigations aimed to improve the insight in the physics related to this isotope effect. Due to the higher atomic mass of deuterium, caesium is removed much more effectively from reservoirs at the walls, resulting in a depletion of these reservoirs and a strongly increased caesium density in the plasma. Additionally, a correlation between the fluxes of charged particles towards the inner ion source surfaces and the co-extracted electrons is identified.

The negative-ion based neutral beam injector for heating and current drive of the ITER plasma (ITER HNB) is under development, at present focusing on the optimization of the full-scale plasma source in the SPIDER test stand. The production of H- or D- ions in the ion source is based on the low work function surfaces obtained by caesium evaporation. This paper describes the caesium conditioning procedure and the corresponding beam performances during the first operation of SPIDER with caesium. Technical solutions to overcome present limitations of the test stand are described. The influence of source parameters on the caesium effectiveness was investigated in short beam pulse operation; with total RF power of 400 kW and filling pressure below 0.4 Pa, and a limited number of extraction apertures, a negative ion current density of about 200 A/m2 was extracted in hydrogen, with beam energy lower than 60 keV. Beam optics and beam uniformity were assessed thanks to the acceleration of isolated ion beamlets. A possible procedure to accelerate a uniform beam was demonstrated at low RF power. The results obtained in this first investigation provided key indications on the operation of one of the largest existing sources of accelerated negative hydrogen-like ions.

The neutral beam heating system for the future international fusion experiment ITER will be based on RF driven ion sources delivering a large (≈12 m) and homogeneous negative hydrogen or deuterium ion beam of several ten Amperes over up to one hour. Such beams have never been produced up to now and a dedicated R&D process is ongoing for more than two decades. An important intermediate step is the size scaling test facility ELISE (Extraction from a Large Ion Source Experiment) with its half-ITER size ion source. Recently, ELISE has fulfilled its first main aim, demonstrating in hydrogen ITER-relevant accelerated negative ion current densities over 1000 s, at the required filling pressure of 0.3 Pa, electron-ion ratio below one and beam homogeneity better than 90 %. The measures identified as essential for achieving such pulses are the introduction of external permanent magnets and internal potential rods as well as a dedicated caesium conditioning technique.

The negative ion source test facility ELISE represents an important step in the European R&D roadmap for the neutral beam injection (NBI) systems of ITER. Its aim is to consolidate the design and to gain early experimental experience with a large and modular Radio Frequency (RF) negative ion source and an ITER like extraction system of the same width but half the height of the ITER source (0.9 × 1 m 2 ). Hand Dbeams can be extracted and accelerated up to 60 kV for 10 s every 150 s out of the continuously operating plasma source.For short plasma pulses (10 s), the extracted negative ion current densities in hydrogen have almost reached the ITER requirement (329 A/m2 H‾, 286 A/m2 D‾). Also the required long pulse source operation up to 1000 s (H-) / 3600 s (D-) could be demonstrated on ELISE with repetitive beam blips, but with reduced current densities. The main limitations are the amount and temporal stability of co-extracted electrons, especially in deuterium operation. This co-extracted electron current has to remain below the extracted ion current to avoid thermal overloading of the extraction grid. Magnetic field configurations, electric potentials of source surfaces close to the extraction system and caesium management are under investigation as tools for source performance optimization. Furthermore RF issues such as heating of source components, RF breakdowns and RF matching have been solved for high power source operation.